JP2018180649A - INFORMATION PROCESSING DEVICE, CONTROL DEVICE, CONTROL PROGRAM, AND RECORDING MEDIUM - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize an information processing device or the like capable of selecting an appropriate control content by grasping that the operating state of equipment is a precursor state of abnormality. An information processing apparatus (200A) has a verbalization processing unit (211) for generating verbalization state information (232a) in which state information related to an operating state of an incinerator (11) is verbalized, and a control candidate identification unit. (212), and the control candidate identification unit (212) refers to the historical verbalization database (232) based on the verbalization state information (232a) regarding the current operating state of the incinerator (11). Identify candidates for control content to be performed on the incinerator (11). [Selection diagram] Fig. 1

Description

  The present invention relates to an information processing apparatus or the like which selects control content based on the state of equipment.

  Conventionally, most of the normal operation of facilities such as plants and waste incineration facilities is automated. On the other hand, at the time of abnormality or a precursor of abnormality, the operator manually performs an operation (manual intervention).

  In such facilities, techniques have been proposed to further increase the efficiency of operation.

  For example, the control device described in Patent Document 1 includes a storage device that stores an operator's operation history and an event history such as an alarm, and an extraction unit that extracts a time-series regularity of operations and events. . The controller predicts the next operation based on the extracted regularity when an operation or an event occurs, and presents the predicted operation to an execution or an operator.

  Moreover, the abnormality prediction system described in Patent Document 2 comprehensively evaluates whether or not an abnormality occurs from the detection values at a plurality of parts constituting the plant by a fuzzy measure / integration, and predicts the occurrence of an abnormality.

Unexamined-Japanese-Patent No. 2006-72408 (March 16, 2006 publication) Unexamined-Japanese-Patent No. 9-228021 (October 31, 1997 publication)

  In the operation of facilities such as plants and waste incineration facilities, manual intervention may be performed if (i) an alarm for notifying any abnormality is issued, or (ii) various sensors for monitoring the operating status of the facility If it is determined that the operator requires an operation (which is a precursor of an abnormality) based on the measured value or the like, the process may be performed. From the viewpoint of the load on each part of the equipment, the efficiency of operation, and the safety, etc., it is required to perform manual intervention before an alarm is issued, that is, at a precursor stage of an abnormality.

  Here, equipment such as a plant includes various control target units, sensors, and the like. Therefore, it is possible to appropriately determine the state of the facility based on various information obtained from a plurality of sensors, and to select an appropriate control target unit based on the determined state and to perform an appropriate operation. It's not easy. Usually, the state determination and manual operation of the equipment are performed based on the experience and intuition of the operator. If the control program can substitute such judgments and operations that require human skill, it is possible to make the operation of the facility more efficient and automated.

  The control device described in Patent Document 1 predicts the next operation of the event based on the regularity extracted in advance when an event such as an alarm notification occurs, and captures a precursor of an abnormality. Is not considered.

  Further, in the anomaly prediction system described in Patent Document 2, when the comprehensive evaluation value obtained by comprehensively evaluating the state of the plant by the fuzzy measure / integration exceeds a predetermined threshold (for example, 0.5), It predicts that an abnormality will occur. That is, this abnormality prediction system predicts an abnormality and performs operation support for the operator's prior abnormality avoidance operation. What kind of operation is actually performed is left to the judgment of the operator, and the above abnormality prediction system does not select an appropriate operation.

  The present invention has been made in view of the above-mentioned conventional problems, and an object thereof is an information processing apparatus capable of grasping that the operation state of equipment is a precursor state of abnormality and selecting appropriate control contents. And so on.

  In order to solve the above problems, an information processing apparatus according to an aspect of the present invention includes a conceptualization processing unit that generates conceptualization information conceptualizing state information related to an operating state of the equipment in an information processing apparatus in a facility performing process control. And a recording processing unit that records control information related to a predetermined type of control content for the equipment and the conceptualization information related to the control information in a database in association with the conceptualization information generated by the conceptualization processing unit And a control candidate identification unit that identifies control content candidates to be performed on the facility by referring to the database based on the information.

  According to the above configuration, in the past operation of the facility, the control information on the control content when the predetermined type of control is performed and the conceptualization information on the condition of the facility at that time are associated with each other and the database It is recorded. Then, in the current operation of the equipment, the conceptualization processing unit generates conceptualization information that conceptualizes the state information of the equipment at the current time, and the control candidate identification unit refers to the database based on the generated conceptualization information.

  The control candidate identifying unit refers to the database, and when conceptualization information on a past operation state of a facility corresponding to or similar to conceptualization information on a current operation state of a facility is present, the control candidate specifying unit relates to the past operation state of the facility Identify control content candidates corresponding to conceptualization information.

  That is, the information processing apparatus should perform the control content executed by the operator in the past operation on the facility at the current time when there is a similar operating condition of the facility in the past based on the current condition of the facility. It can be selected as the control content. Here, in the operation state of the facility, when the state in which the alarm is issued is an abnormal state, the timing at which the operator performs control (for example, manual operation) may be a precursor state of an abnormality. Also, the control content may be appropriate based on the experience and intuition of the operator that is difficult to quantify. Appropriate operations at such appropriate timing are recorded in the database as the conceptualization information and control content, and the control candidate identifying unit determines the current operating condition of the facility with reference to the database and is appropriate. Operations can be identified.

  Therefore, it is possible to realize an information processing apparatus capable of selecting an appropriate control content by grasping that the operation state of the facility is a precursor state of abnormality.

  Further, in the information processing apparatus according to one aspect of the present invention, the conceptualization processing unit may generate, as the conceptualization information, linguistic data representing the state information in a language.

  According to the above configuration, the state information of the facility is verbalized and stored as verbalized data in the database. Therefore, while being able to verbalize and express the various numerical information regarding the driving | running state of an installation, the tendency of the change of the time series data regarding the operating state of an installation can be expressed easily. Then, such verbalization can be performed to reflect the sense of the operator in the operation of the facility, and the ambiguity of the language can be used to express the operating state of the facility.

  As a result, by referring to the database on the basis of verbalization data that verbalizes the current equipment state information, it is possible to inquire about similar past operation states. Then, it is possible to refer to the appropriate control contents in the precursory state of the abnormality performed by the operator in the past operation. Therefore, it can be grasped that the operating condition of the equipment is a precursor condition of abnormality, and appropriate control contents can be selected.

  Also, language similarities can be exploited to reference similar past driving conditions. Therefore, even if there is no past operation state that matches the verbalization data in which the current equipment state information is verbalized, it is possible to identify candidates for control content by applying similar ones with similar numbers. .

  Furthermore, the control candidate identification unit can easily refer to the database. Further, the amount of information recorded in the database can be reduced, and the user can easily grasp and edit the recorded information.

  The information processing apparatus according to one aspect of the present invention may further include an evaluation unit that specifies a plurality of candidates for the control content, and ranks the candidates based on a predetermined criterion. .

  According to the above configuration, the user can select an appropriate control content from among the plurality of specified control content candidates with reference to the ranking. The evaluator can also correct the ranking based on the result selected by the user. Therefore, the control candidate identification unit can identify appropriate control content based on the corrected ranking.

  In the information processing apparatus according to one aspect of the present invention, the conceptualization processing unit may generate, as the conceptualization information, type conceptualization information on each of a plurality of types of state information indicating an operation state of the facility.

  According to the above configuration, it is possible to more appropriately conceptualize the state information regarding the operating state of the equipment.

  In order to solve the above-mentioned subject, the control device in one mode of the present invention is provided with the above-mentioned information processor, a selection part which makes a user select control contents, and a control processing part which performs the selected control contents. ing.

  According to the above-described configuration, it is possible to realize a control device that selects an appropriate control content and executes the selected control content in a precursor condition of an abnormality in the operation of the facility.

  Moreover, in order to solve said subject, the control apparatus in 1 aspect of this invention is provided with the said information processing apparatus and the control processing part which performs the control content identified by the said control candidate identification part.

  According to the above configuration, there is no need for the user to select the control content. Therefore, the control device capable of achieving further automatic operation of the operation of the facility can be realized.

  In order to solve the above problems, an information processing method according to an aspect of the present invention generates conceptualization information conceptualizing state information related to the operating state of the equipment, in a control method of an information processing apparatus in equipment performing process control. A recording processing step of correlating the conceptualization processing step, control information related to a predetermined type of control content for the equipment, and the conceptualization information related to the control information, and recording the same in a database; And a control candidate identification step of identifying control content candidates to be performed on the facility by referring to the database based on the conceptualization information.

  According to the above configuration, the same effect as that of the above-described information processing apparatus can be obtained.

  The information processing apparatus according to each aspect of the present invention may be realized by a computer. In this case, the computer is operated as each unit (software element) included in the information processing apparatus, and the information processing apparatus is operated by the computer. A control program of an information processing apparatus to be realized and a computer readable recording medium recording the same also fall within the scope of the present invention.

  According to an aspect of the present invention, it is possible to realize an information processing apparatus or the like that can select an appropriate control content by grasping that the operation state of a facility is a precursor state of an abnormality.

It is a block diagram which shows schematic structure of the information processing apparatus in Embodiment 1 of this invention. BRIEF DESCRIPTION OF THE DRAWINGS It is a block diagram which shows the outline | summary of the driving assistance system containing the information processing apparatus in Embodiment 1 of this invention. It is the time-series data of the operating quantity of the control object part provided in the waste heat boiler as an example of process data. It is a figure for demonstrating the method to verbalize the change tendency of time series data. It is a flowchart figure which shows the flow of the processing which constructs | assembles a historical language-ized database using the existing control historical data. (A) is a figure for demonstrating the normal data and precursor data in process data, (b) is a table | surface which shows an example of the result of having verbalized the said precursor data part, and the membership value of each word is there. It is a figure for demonstrating neural network processing. It is a flowchart figure which shows the flow of the information processing by the said information processing apparatus. It is a flowchart figure which shows the flow of the process which the recording process part of the said information processing apparatus performs. It is a block diagram which shows schematic structure of the information processing apparatus in Embodiment 2 of this invention. It is a figure for demonstrating generation | occurrence | production of linguistic state integrated information based on process data. It is a flowchart figure which shows the flow of the processing which constructs | assembles a historical language-ized database using the existing control historical data. It is a flowchart figure which shows the flow of the information processing by the said information processing apparatus.

Embodiment 1
One embodiment of the present invention will be described below with reference to FIGS.

  In the present embodiment, most of the normal operation is automated, and a waste incineration facility is illustrated as a facility provided with various sensors and various control target parts, and the information processing applied to the operation of the waste incineration facility The apparatus will be described. Note that the information processing apparatus according to one aspect of the present invention is not necessarily limited thereto, and, for example, other waste processing facilities, plants (plants), and power plants, etc., which are automated, can be operated at normal times. Most are automated and can be applied to equipment where the operator decides to perform manual control if it determines that manual intervention is required.

  In order to facilitate understanding of the information processing device in one aspect of the present invention, first, the driving support system 1 including the information processing device 200A will be schematically described using FIG.

<Driving support system 1>
FIG. 2 is a block diagram showing an outline of the driving support system 1 including the information processing device 200A in the present embodiment. The driving support system 1 includes a waste incineration facility (a facility that performs process control) 10 and a management center 20, which are communicably connected via a communication network 30.

  The driving support system 1 is an example, and the configuration of the driving support system including the information processing device 200A is not limited to this. For example, the information processing apparatus 200A may be provided in the refuse incineration facility 10, and the system may be completed within the facility. In addition, the driving support system according to one aspect of the present invention may include a data center communicably connected to the communication network 30, and various data described later may be stored and accumulated in the data center. May be

(Waste incineration facility 10)
The waste incineration facility 10 is provided with an incinerator 11 for burning waste and a central control room 12 for controlling it.

  The incinerator 11 is, for example, a stoker furnace including a stoker constituted by a grate that moves back and forth, and includes a large number of sensors 13 and a large number of control target parts 14. Each sensor 13 and each control target unit 14 are connected to the system bus 15, respectively. The sensor 13 and the control target unit 14 are of various types, and thus the detailed description will be omitted for convenience of the description.

  The central control room 12 is provided with an operation control device 120, and an operator 121 who operates the operation control device 120 as needed stands by.

  The operation control device 120 receives information from the plurality of sensors 13 through the system bus 15 and controls the plurality of control target units 14 respectively. The operation control device 120 performs the automatic operation of the incinerator 11 when the incinerator 11 is in the normal operation state. The operation control device 120 is provided with a large number of monitors and consoles, and the operator 121 monitors the operation state based on the display of these monitors, and when it is determined that it is an abnormal state or a precursor state of an abnormality, manual operation is Do.

  In general, the waste incineration facility 10 also includes other systems (not shown) such as, for example, power generation, heat utilization, and water circulation, and these systems also include various sensors and control target units. Each of these systems is not independent of one another, and the operation on one system can affect the operating state of the other system. The operator 121 monitors and manually operates the operating state of each of these systems based on various information. There are various parameters in the operation of equipment such as the waste incineration facility 10, and it is difficult to properly determine the operation state of the equipment and select an appropriate operation. Here, since it is the incinerator 11 that the monitoring and control of the operation state is most important in the waste incineration facility 10, in the present embodiment, automatic control of the incinerator 11 is carried out to simplify the explanation. Main target. The information processing apparatus in one aspect of the present invention may control the entire waste incineration facility 10 as a control target.

  In addition, the waste incineration facility 10 may be various types of facilities conventionally used. The incinerator 11 may be a fluidized bed type, a gasification and melting type, or any other type.

(Management center 20)
The management center 20 includes an information processing apparatus 200A, and an operator 21 operating the information processing apparatus 200A is on standby.

  The information processing apparatus 200A includes a control unit 210, a recording processing unit 220, and a storage unit 230. The details of these units will be described later. The information processing device 200A receives process data from the operation control device 120 via the communication network 30. The process data is various time-series data related to the operating state of the incinerator 11 that the operation control device 120 receives from the sensors 13 through the system bus 15. That is, the process data is time-series data indicating the state of the incinerator 11, and a plurality of process data exist corresponding to the number of sensors 13.

(Schematic explanation of driving support by the system)
Conventionally, combustion control of refuse in the incinerator 11 is performed by automatic combustion control (ACC). Generally, in the automatic combustion control, (i) the amount of steam generated in the waste heat boiler provided above the incinerator 11, (ii) the gas temperature of the combustion chamber which burns while transporting waste by stokers, and (iii) exhaust gas The amount of dust supply, the amount of combustion air, etc. are computer controlled based on the value of the component analysis value of, etc. This makes the combustion of waste uniform and stable.

  However, the control of waste combustion can be complicated due to various factors, such as various wastes usually mixed with each other, moisture content being high, high in metal content, high in flammable plastic etc. Therefore, combustion control by a computer may be difficult. In the automatic combustion control, for example, an alarm is issued when a measurement value of a certain sensor exceeds or falls below a predetermined threshold, and a manual operation by a person is requested.

  On the other hand, the operator 121 standing by in the central control room 12 determines the state of the incinerator 11 based on the information from the plurality of sensors 13 and the information from the monitor that displays the image of the combustion chamber.

  Then, the operator 121 performs a manual operation (manual intervention) on the operation control device 120 when the alarm is issued or when it is determined that the alarm is a precursor of an abnormality. For example, when the component of the waste suddenly changes and the state of burning of the waste changes, manual intervention may be performed. In addition, even if a warning is issued, the operation is not necessarily required, and there may be a case where the component of the haiku gas instantaneously increases. In this case, the operator 121 may decide not to perform the manual intervention based on experience and intuition.

{Process data}
Here, various process data indicating the state of the incinerator 11 have time-series fluctuations. The operator looks at the state of the change and determines whether or not it is a sign of anomalous. An example of this process data will be described with reference to FIG. FIG. 3: is time-sequential data of the operating quantity of the control object part provided in the waste-heat boiler as an example of process data. The horizontal axis indicates time (in minutes), and the vertical axis indicates arbitrary units, and indicates changes in time-series values. This operation amount is a mixture of the automatic operation by the operation control device 120 and the manual operation. Here, the "operation amount" does not mean a difference between before and after performing a certain operation, but means, for example, an output value to a control target unit.

  As shown in FIG. 3, the data of the operation amount acquired at predetermined time intervals is time-series data which changes by automatic operation control and manual operation based on the difference between the set value and the measurement value. Become. Further, based on the operating state of the incinerator 11 and the predetermined set value, the limit value of the operation amount slightly fluctuates. The limit value is indicated by a broken line in the figure. This operation amount is changed automatically based on the measurement value of the sensor and manually by the operator 121 so that the amount of generated steam in the waste heat boiler becomes stable (close to the set value).

  In the figure, in the region A1, the operation amount increases rapidly and reaches the limit value, but the operation amount decreases immediately after that. Such a condition does not correspond to an abnormal condition. When the operator 121 has a high level of skill, it can be said that this is a case where the information from various sensors is comprehensively judged and the manual intervention is not performed.

  In the region A2, the manipulated variable generally increases as a trend, and then reaches the limit value. Such a change of state is caused, for example, by the presence of inflammable waste. And in area | region A3, the state whose operation amount is a limit value continues, and such a state corresponds to an abnormal state.

  Thereafter, in the region A4, the amount of operation decreased as a general trend. Such a change in state is caused, for example, by the increase of combustible waste in the combustion chamber. Further, in the region A5, it can be said that the operation amount does not reach the limit value, and is in a stable state (normal operation state). Even in the stable state, it can be seen that there is a considerable increase and decrease in the amount of operation.

  In such process data, for example, it is very difficult to perform an appropriate operation (manual intervention) at the stage such as (i) a stage such as the area A2, judging that it is a precursor state of an anomaly and at that stage I understand. And (ii) determining that the operation amount as in the region A1 rapidly increases, that is, at the time when the sensor value (measurement value) changes rapidly, and not an abnormal state, and (iii) in the region A5 It is also difficult to appropriately determine, for example, that it does not correspond to the precursor condition of abnormality at the time when the operation amount is increasing.

  Here, when the proficiency level of the operator 121 is high, the operator 121 appropriately determines the above (i) to (iii) while referring also to information from various other sensors, and performs an appropriate operation. be able to. In addition, this “proper operation” includes, for example, an element such as at what timing, at what time, and at what rate, a certain control value is changed. On the other hand, when the experience of the operator 121 is insufficient, the operator 121 may determine the operation state inappropriately or perform an inappropriate operation.

  It is difficult to quantify the skills possessed by the highly skilled operator 121, such as the timing of the determination of the operation state and the selection and operation ratio of the manual operation. Therefore, the computer control such as conventional automatic combustion control appropriately determines the state of the incinerator 11 based on a predetermined rule such as a threshold and selects an appropriate operation (see It is very difficult to automate.

{Driving support in one aspect of the present invention}
Here, there are many intervention logs (manual operation history data) as follows as a result of the past operation in the refuse incineration facility 10. That is, various process data of the incinerator 11 from a little before the time when the operator 121 performs the manual control, contents of the manual control, and various kinds of the incinerator 11 from the time after the manual control to some time If a series combination of process data is referred to as "intervention log", a large number of the intervention logs are accumulated and stored. The “various process data” includes time-series data obtained from various control target units 14 and sensors 13.

  It is difficult to record and use many accumulated past intervention logs (numerical information) as described above in a database as they are. This is not realistic because the amount of information in the database becomes enormous and the processing time for extracting appropriate information with reference to such a bundle of process data becomes too long. It is. Also, numerical data is inherently difficult to refer to. This is because the process data has numerical value fluctuations, and the tendency of time series change may be more important than the absolute value of numerical values, and so on.

  Therefore, the driving support provided by the information processing device 200A in the present embodiment is roughly as follows. That is, the information processing apparatus 200A conceptualizes the operation state (various process data) of the incinerator 11 for each of the many past accumulated intervention logs as described above, and generates verbalized information (verbalized data). At the same time, the verbalization information and the control content are correlated and recorded in the database. Here, since the process data is time-series data, it is preferable to reflect the tendency of temporal change in its verbalization. The method of conceptualizing the intervention log is not limited to verbalization.

  For example, conceptualization such as verbalization makes it possible to practically refer to the information recorded in the database. Such verbalization can be performed to express the operating state reflecting the sense of the operator in the operation of the facility. The ambiguity of the language can be used to express the operating condition of the equipment. In addition, verbalization can save the amount of data stored in the database. In addition, in the case of modifying the database later, the modification can be made relatively easy.

  Then, the current operation state (various process data) of the incinerator 11 is conceptualized in the same manner as the verbalization information, and the incinerator 11 is referred to the past verbalized intervention log corresponding to the verbalization information. Identify candidates for control content to be performed. Thereby, the information processing apparatus 200A appropriately uses the information of the determination and operation (the past manual intervention operation) which the operator 121 has made based on the experience and intuition in the past, and is appropriate for the current operation state of the incinerator 11. Control content can be selected.

  The user can appropriately set the width of the time “from a little before” and “to a little after” in the intervention log described above. Preferably, the width of the time may be determined based on the trend of temporal change, which will be described in detail later.

  Although details will be described later, the information processing apparatus 200A selects control contents by referring to verbalized information such as document data such as various specifications and academic papers, as well as past history data. May be Further, the database may not be built in advance using historical data in the past, and the information processing device 200A may learn an appropriate state determination and operation while operating the driving support system 1.

<Information processing device 200A>
The information processing apparatus 200A included in the driving support system 1 according to one aspect of the present invention, the outline of which has been described above, will be described in detail below with reference to FIG. FIG. 1 is a block diagram showing the main configuration of an information processing apparatus 200A in the present embodiment.

  As shown in FIG. 1, the information processing apparatus 200A includes a control unit 210, a recording processing unit 220, a storage unit (database) 230, a display unit 240, and an input unit 250. The information processing apparatus 200A is communicably connected to the operation control apparatus 120 as described above.

  The operation control device 120 includes an operation control unit 120a, a display unit 120b, an input unit 120c, and a data storage unit 120d. The operation control unit 120 a is connected to the above-described system bus 15 to perform an automatic operation of the incinerator 11. The display unit 120 b is a plurality of monitors that display measurement values and the like of the various sensors 13. The input unit 120c receives an input such as a manual operation of the operator 121 (see FIG. 2). The data storage unit 120 d stores various data received from the incinerator 11 and the contents of the manual operation of the operator 121. The operation control device 120 transmits various process data received from the incinerator 11 and manual intervention information indicating that the operator 121 has manually intervened to the information processing device 200A.

  The control unit 210 of the information processing device 200A includes a linguistic processing unit (conceptification processing unit) 211 and a control candidate identification unit 212. Further, the control candidate identification unit 212 includes a determination unit 212 a and an evaluation unit 212 b. Then, the storage unit 230 includes languageization rule data 231, a history languageization database 232, and a process data storage unit 233. The verbalization state information 232a, the control information 232b, and the evaluation information 232c are stored in the history linguisticization database 232 in association with each other. A detailed description of each of these units will be given later together with a description of the operation of the information processing apparatus 200A, but will be schematically described below.

  The control unit 210 integrally controls the respective members of the information processing apparatus 200A. The control unit 210 receives various process data (state information) transmitted from the operation control device 120. Based on the various process data received, the verbalization processing unit 211 uses the verbalization rule data 231 to verbalize (conceptify) the operation state of the incinerator 11, and the verbalization state information on the current operation state ( Generate conceptualization information, verbalization data).

  Then, using the verbalization state information on the current driving state generated by the verbalization processing unit 211, the determination unit 212a of the control candidate identification unit 212 first determines whether the operating state of the incinerator 11 is a precursor state of abnormality It is determined whether or not. In the present embodiment, the determination unit 212a is a neural network processing unit that determines the operating state of the incinerator 11 by performing neural network processing. The neural network processing will be described later.

  If the determination unit 212a determines that the operating condition of the incinerator 11 is a precursor condition of an abnormality, the control candidate identification unit 212 refers to the history linguisticization database 232 based on the verbalization condition information used for the determination. And identify appropriate control content candidates. Specifically, the operating state of the incinerator 11 that matches or is similar to the verbalization state information on the current operating state used for the determination is retrieved from the verbalization state information 232a of the history linguisticization database 232. And the control content matched with the applicable linguistic state information is specified as a control candidate.

  The control content candidates specified by the control candidate specifying unit 212 are displayed on the display unit 240. In other words, based on the operating condition of the incinerator 11, candidates for control content to be performed are proposed to the operator 21 (see FIG. 2). The operator 21 selects an appropriate control content from among the candidates displayed on the display unit 240 using the input unit 250.

  The control unit 210 transmits an intervention control instruction to the operation control apparatus 120 based on the control content selected by the operator 21. The operation control device 120 performs intervention control on the incinerator 11 based on the received intervention control instruction.

  The process data storage unit 233 may store a part of the existing control history data accumulated in the past operation, and may store the process data received from the operation control device 120.

  When the control unit 210 receives the manual intervention information from the operation control device 120, the recording processing unit 220 associates the control content, the operation status, and the evaluation information described later, and stores them in the history linguisticization database 232. Here, the recording processing unit 220 may be configured to perform the verbalization processing using the verbalization processing unit 211 of the control unit 210.

  Further, the recording processing unit 220 may use data stored in the process data storage unit 233. Process data stored in the process data storage unit 233 is stored in the process data storage unit 233. The process data storage unit 233 is not an essential component, and the recording processing unit 220 may be configured to acquire data from the data storage unit 120 d of the operation control apparatus 120 or the like.

  Here, the information processing apparatus 200A of the present embodiment includes a display unit 240 and an input unit (selection unit) 250 that allows the user to select control content, and a control unit (control processing unit) that executes the selected control content. It can be said that it is a control device provided with 210. In addition, the information processing apparatus 200A may not include the input unit 250. In this case, the information processing apparatus 200A can support the operation of the incinerator 11 by displaying candidates for appropriate control content for the user (the operator 21 or the operator 121).

(Language of driving condition)
Here, the verbalization processing based on the received process data performed by the verbalization processing unit 211 described above will be described with reference to FIG. FIG. 4 is a diagram for explaining a method of verbalizing the tendency of change of time series data.

  The verbalization processing unit 211 receives multiple types of process data for each of the multiple sensors 13 of the incinerator 11. The plurality of sensors 13 may include an image sensor. For example, as information from the sensor 13, image data obtained by capturing an image of waste combustion may be included.

  Although the verbalization of process data is schematically described below, the method of verbalization is not limited to this. The type of process data changes depending on what sensor 13 is provided in the incinerator 11. Also, which process data should be verbalized and how to set the rules of verbalization may be appropriately set by the user. Such a set rule becomes the verbalization rule data 231.

  In addition, since it is possible to apply the well-known technique to the algorithm used for linguisticization, it abbreviate | omits about the description.

(1) Simple language conversion of measurement value It can be simply verbalized using the measurement value of each sensor at each time which forms process data. For example, it can be verbalized such as “in-furnace temperature is 950 ° C.”, “feed water flow rate is 13 t / h”, and “dust collection speed is 3.3 m / h”. Such verbalization is relatively easy, but on the other hand, it is not suitable for text matching when referring to a database later. This is because, for example, when the data stored in the database is 500 ° C., if the operating condition of the equipment is 501 ° C., the matching does not occur.

(2) Process state language conversion The operating state of the incinerator 11, for example, "Steam weight is stable", "CO peak is generated", "Grate temperature is slightly unstable" Evaluate measured values based on a preset range and verbalize from various angles, such as "the steam temperature is abnormally high" and "the combustion grate speed has reached the upper limit" it can. Here, “the steam weight is stable” may be determined as such when, for example, the fluctuation of the vapor weight is within ± 2.5%, the range of fluctuation is May be set appropriately. A numerical range (variation range) for making a determination may be appropriately set for each of various process data.

  In addition, the time range for performing such determination may be set as appropriate. The set linguistic rule is the linguistic rule data 231. The verbalization rule data 231 created is used consistently in the driving support system 1. When changing the language setting rule set beforehand during operation of the driving assistance system 1, the history language-ized database 232 will be rebuilt.

  Note that state languageization of such processes may be performed using fuzzy measures / integrals or other known methods.

(3) Language conversion of the diagnosis result For example, the operation state of the facility may be diagnosed based on the result of language conversion as described in (1) and (2) above, and a preset rule. For example, it can be verbalized as "there is a considerable margin in incinerator load", "there is no margin in steam amount control", and "the waste quality is stable".

(4) The verbalization of the determination result by the image sensor With regard to the image data from the image sensor as the sensor 13 provided in the combustion chamber of the incinerator 11, the waste combustion state is determined by using a known image analysis algorithm. Can. The determination result can be verbalized as, for example, “35% burn-out, 50% unburned mass, 15% normal”. Based on the verbalized information on the combustion state, the operation state of the facility may be further conceptualized and verbalized, for example, “the quality of waste is a little bad”.

(5) Linguization of time-series change It is preferable to cluster time-series data to verbalize a trend. This trend languageization can be performed, for example, as follows. That is, as shown in FIG. 4, with respect to time series data as certain process data, first, the time series data is divided into appropriate clusters at portions having similar trends. This division can be performed based on the determination of the degree of similarity between clusters. The determination of the degree of similarity between clusters can be performed using, for example, three values of the value of the cluster, the amount of change of the cluster, and the degree of vibration of the cluster. An algorithm for dividing time series data into appropriate clusters can use a known algorithm.

  Next, the features of the clustered time series data are extracted and verbalized. Thereby, for example, the tendency of certain clustered time series data can be verbalized as “30% convex, 10% concave, 60% Δ”. Based on this result, for example, it may be symbolized as “type B”. Also, it can be verbalized as, for example, "large increase" or "small increase". This trend linguisticization can also be verbalized by categories such as the general tendency of clusters, and the pre-, mid-, and late trends.

  Also, a figure (for example, a bar graph) may be created and stored in the database based on data verbalized to the tendency of time series data. For example, for a cluster in which “medium rise” is a trend of 80% and “concave downward” is a trend of 20%, bar graphs can be created with various evaluation items as the horizontal axis and the vertical axis as the ratio. In addition, as a figure, it is only necessary to be able to express the tendency (such as size) of time-series data indicating changes in numerical values and the tendency (area, color, etc.) of time-series data indicating changes in images. The aspect is not particularly limited.

  The language conversion rules as described above are stored in the storage unit 230 as language conversion rule data 231. Note that the above is an example, and the rule of verbalization can be appropriately set in advance.

(History Linguization Database 232)
Next, with reference to FIG. 5, the flow of processing for constructing the history languageization database 232 using the existing control history data accumulated in the past operation of the refuse incineration facility 10 will be described. The control history data is data in which process data in the past operation of the refuse incineration facility 10 is stored, and includes at least the aforementioned intervention log. FIG. 5 is a flow chart diagram showing a flow of processing for constructing the history languageization database 232 using existing control history data.

  As shown in FIG. 5, first, the stored existing control history data is prepared (step 1: abbreviated as S1 below). The information processing apparatus 200A reads this control history data.

  The control unit 210 extracts information on one intervention log from the information on the plurality of intervention logs included in the control history data (S2).

  The verbalization processing unit 211 performs verbalization processing on the extracted intervention log as follows. That is, the verbalization processing unit 211 uses the verbalization rule data 231 to verbalize the operating state based on various process data of the waste incineration facility 10 from the time of control in the past to the first predetermined time ago. Languageization status information (S3: conceptualization processing step). Furthermore, the linguistic processing unit 211 may verbalize each of the various process data individually, and generate plural types of type linguistic status information (type conceptualization information) as linguistic status information (conceptual information). Good.

  In addition, the verbalization processing unit 211 verbalizes the operating state based on various process data of the waste incineration facility 10 from the time when control is performed in the past to the second predetermined time later and sets it as verbalization control result information ( S4). This verbalization control result information may also include a plurality of types of type verbalization control result information.

  The first predetermined time and the second predetermined time may be set as appropriate. Also, different first predetermined times may be set for each of various process data. Similarly, different second predetermined times may be set for each of various process data. The first predetermined time and the second predetermined time may be appropriately set in accordance with the characteristics of each process data.

  After that, the recording processing unit 220 associates the verbalization state information 232a with the verbalization control result information and the control content, and records them in the history linguisticization database 232 (S5: recording processing step). Here, the control content and the verbalization control result information are combined to form control information 232 b.

  The processes in S2 to S5 described above can be performed for each of a plurality of intervention logs included in the control history data, and the history verbalization database 232 can be constructed.

  As described above, the history linguisticization database 232 can be constructed with the unified linguistic rule. In such verbalization, the sense of the operator 121 is reflected on the variation range of the measurement value of the sensor 13, the manner of the variation, and the region where the measurement value is between the normal time and the abnormal time, etc. Can be verbalized. In addition, by verbalization, it is possible to refer including the degree of similarity in consideration of the similarity of words.

  In the present embodiment, the information processing apparatus 200A is configured to perform the verbalization processing based on the control history data to construct the history linguisticization database 232, but in one aspect of the present invention, the information processing is performed. The history languageization database 232 may be constructed by a languageization apparatus provided outside the apparatus 200A. The constructed history verbalization database 232 may be transferred to the storage unit of the information processing apparatus 200A.

  In addition, the history linguisticization database 232 may include information on speech and conversation of the operator 121 as the verbalization status information 232a. The operating condition of the installation may be something like a tone that does not appear clearly in the measurements of the sensor 13. As a result, the verbalization state information 232a can be made more reflecting the sense of the operator 121.

(Evaluation part 212b)
Here, the evaluation unit 212 b will be described in relation to the construction of the history linguisticization database 232. As described above, as a result of constructing the history languageization database 232 based on the control history data in the past operation of the waste incineration facility 10, the correspondence between the languageization state information 232a and the control information 232b is as follows: Can occur.

  That is, a plurality of control information 232b may be associated with certain verbalization state information 232a. This is because in a certain operating state of the refuse incineration facility 10, the control content input to the operation control device 120 is not always the same, and may differ from time to time.

  In particular, when the operator is different, the determination of the operation state and the content of the manual operation based on the determination may differ depending on factors such as the operator's skill level. Therefore, with respect to certain verbalization state information 232a, both control information related to more effective control content and control information related to control content with relatively small effect may be associated.

  Therefore, the evaluation unit 212b ranks a plurality of pieces of control information 232b associated with certain verbalization state information 232a based on a predetermined criterion. The ranking based on the predetermined criteria can be performed, for example, using the verbalization control result information. As a result of performing a certain control content, it is possible to evaluate the control content based on the time required for the operating state to return to a stable state. Further, it is preferable that the evaluation unit 212b changes the weighting according to the state and performs ranking, since the matters to be prioritized are different (the priorities are different) according to the specific state of each intervention log. .

  Then, information about the result of evaluation by the evaluation unit 212 b is stored in the history linguisticization database 232 as evaluation information 232 c. That is, the verbalization state information 232a, the control information 232b, and the evaluation information 232c are stored in the history linguisticization database 232 in association with each other.

  Note that this evaluation information 232c may be appropriately adjusted by the user. Further, as described later, a plurality of operation candidates may be presented to the operator during operation of the facility, and the evaluation value of the operation selected by the operator may be increased.

(Determination unit 212a)
Next, with reference to FIG. 6 and FIG. 7, the operation performed by the determination unit 212a that determines whether or not the operation state is normal in the current operation of the refuse incineration facility 10 will be described. (A) of FIG. 6 is a figure for demonstrating the normal data and precursor data in process data. FIG. 6B is a table showing the results of verbalizing the precursor data portion and the membership value of each word. FIG. 7 is a diagram for describing neural network processing.

  In the process data shown in FIG. 6A, area A11 is normal data, and area A12 is precursor data indicating a precursor condition of abnormality. An example of a verbalized trend of the change of the cluster is shown in (b) of FIG. 6 for this precursor data portion.

  Here, as shown in (b) of FIG. 6, verbalization is performed in five categories, such as the general tendency, the period of the local feature, the value of the local feature, the position of the late stage, and the vibration of each period. It is a rule. Furthermore, as expressions used in each category, it is defined as follows. The overall trend is 11 types (large increase, downward convex, etc.). The period of the local features is 3 types (early, middle, late). The values of local features are 7 types (moderate increase, slight decrease, etc.). There are four types of late positions (nearly zero, low, medium, high). The vibration in each period is 16 types for each of the early period, middle period, and late period (4 types of expression of standard deviation and 4 types of expression of frequency). That is, the total number of types of words is 11 + 3 + 7 + 4 + 16 × 3 = 73 types.

  As shown in FIG. 7, the determination unit 212a can perform neural network processing to determine the operating state of the incinerator 11. A known technique can be used as such neural network processing, and the detailed description is omitted. Generally speaking, membership values of the above-mentioned 73 kinds of words are input to the input layer. Then, neural network processing is performed to obtain an output value. This output value can be used to determine whether the operating condition is normal.

  A neural network that performs such processing can be constructed in advance by machine learning. For example, when the output value is (1, 0), it can be set as precursor data and when it is (0, 1), it can be supervised learning as normal data. In FIG. 7, since the resulting output value is (0.548, 0.452), in this case, it is determined to be precursor data. By preparing a large number of normal data and precursor data and performing supervised learning, the accuracy of determination by the neural network is improved.

  Note that the information processing apparatus 200A may not include such a determination unit 212a. In that case, the verbalization processing unit 211 performs verbalization at a predetermined time interval with respect to the current operation state of the incinerator 11, and the control candidate identification unit 212 determines, based on the verbalized information, a history verbalization database Reference numeral 232 may be referred to at predetermined time intervals.

  Details of the operation of the control candidate identification unit 212 including the determination unit 212a and the evaluation unit 212b will be described in conjunction with the following description of the flow of information processing executed by the information processing apparatus 200A.

<Operation Selection Process Performed by Information Processing Device 200A>
The flow of information processing by the information processing apparatus 200A will be described with reference to FIG. FIG. 8 is a flowchart showing the flow of information processing by the information processing apparatus 200A.

  As shown in FIG. 8, first, in a normal state, the refuse incineration facility 10 is automatically operated, and the information processing apparatus 200A receives process data from the operation control apparatus 120 (S11).

  If the predetermined time has not elapsed (NO in S12), the process of S11 is repeated. When the predetermined time has elapsed (YES in S12), the verbalization processing unit 211 verbalizes the current operation state of the incinerator 11 using various process data, and generates verbalization state information (S13) .

  Based on the verbalization state information generated by the verbalization processing unit 211, the determination unit 212a determines whether the operation state of the incinerator 11 is a precursor state of abnormality. Here, the determination unit 212a according to the present embodiment determines the operating state of the incinerator 11 using a neural network (S14).

  If it is determined that the operating state of the incinerator 11 is not a precursor of abnormality (NO in S15), the processing from S11 is repeated again. If it is determined that the operation state of the incinerator 11 is a precursor state of abnormality (YES in S15), the control candidate identification unit 212 determines the history linguistic database 232 based on the verbalization status information generated in S13. Refer to (control candidate identification step). Here, the method by which the control candidate identification unit 212 refers to the history linguisticization database 232 is not particularly limited, but may be text matching as an example. In addition, graphs or image data may be stored in the history verbalization database 232, and in this case, pattern matching may be performed. An algorithm for performing such a search may be a known one and is not particularly limited.

  When there is no linguistic state information 232a corresponding to the history linguisticization database 232 (NO in S16), the control candidate identification unit 212 notifies the display unit 240 or the display unit 120b that the operation by the operator is necessary. ((A) of FIG. 8). When there is the corresponding verbalization state information (YES in S16), the control candidate identification unit 212 displays the control content candidates on the display unit 240 in the order of evaluation based on the corresponding verbalization state information.

  Here, the "corresponding verbalization state information" may be identical or similar. The control candidate identification unit 212 may refer to the history linguisticization database 232 based on the degree of language similarity. In addition, when the verbalization state information generated in S13 includes a plurality of types of verbalization state information, the control candidate identification unit 212 determines a type that matches or is similar to the combination of the plurality of types of verbalization state information generated in S13. A combination of verbalization state information may be specified from the history languageization database 232, and the control content candidate corresponding to the combination may be specified.

  In addition, the control candidate identification unit 212 may identify candidates for control content based on the verbalization state information 232 a having a large number of similarities by referring to the history linguisticization database 232.

  Further, it is preferable that the control candidate identification unit 212 presents candidates for the control content on the display unit 240, and display reasons for selection and evaluation of the candidates. Thereby, the operator 21 can select an appropriate control content in consideration of the reason.

  Next, when the operator 21 selects the control content using the input unit 250 from among the control content candidates displayed on the display unit 240 (YES in S18), the control unit 210 is selected The intervention control instruction is transmitted to the operation control device 120 based on the control content. That is, control part 210 performs intervention control to automatic operation control in operation control device 120 (S19). Thereafter, the evaluation unit 212b increases the evaluation of the control information related to the selected control content with respect to the plurality of control information associated with the verbalization state information 232a in the history language conversion database 232 (S20).

  On the other hand, when the operator 21 does not select the control content using the input unit 250 from among the control content candidates displayed on the display unit 240 in S18 (NO in S18), the display unit 240 or The display unit 120b is notified that an operation by the operator is necessary ((A) in FIG. 8).

(Information addition processing of history language database)
The flow of processing performed by the recording processing unit 220 of the information processing device 200A when the manual operation by the operator is performed will be described with reference to FIG. FIG. 9 is a flowchart showing the flow of processing performed by the recording processing unit 220 of the information processing apparatus 200A.

  When there is no corresponding linguistic state information in S16 (see FIG. 8) (NO in S16), or when the operator 21 does not select the control content using the input unit 250 (NO in S18), As shown in FIG. 9, the display unit 240 or the display unit 120b displays that the manual operation by the operator is requested (S31).

  When the operator manually executes the operation (YES in S32), process data from the time when the operation is performed to the second predetermined time is acquired. Here, the process data from the operation control device 120 is transmitted to and stored in the process data storage unit 233 (see FIG. 1). The recording processing unit 220 acquires process data from the process data storage unit 233. Alternatively, the process data may be acquired from the data storage unit 120d of the operation control device 120.

  The recording processing unit 220 verbalizes the acquired process data to obtain linguistic control result information (S33). After that, the recording processing unit 220 associates the verbalization state information, the control content, and the verbalization control result information, and records them in the history linguisticization database 232 (S34).

  If the operator does not manually execute the operation in S32, the automatic operation is continued, the operation is stopped, or an alarm is issued (S35).

(Main effect)
Generally, in the control of the incinerator 11, it is difficult to read from the control history data in the past as to which of the sensors 13 the operator 121 has made the determination and performed the operation. In addition, the operation of the operator 121 is not simple, and may slightly change the operation amount of one control target unit 14 and then further change the operation amount of another control target unit 14. Furthermore, the change of the operation amount has an element of time, and may be changed at once or changed in small steps. The reason for such operations, ie human thinking and judgment, can not be data output. That is, it is difficult to promote automatic operation of the incinerator 11 by utilizing the past control history data, because the reason for the manual operation in the past control history data is not clear.

  Therefore, in the information processing apparatus 200A according to one aspect of the present invention, the operation state of the incinerator 11 in the control history data in the past is conceptualized (verbalized) according to a unified rule based on a predetermined rule. By verbalization, it is possible to express various measurements of the sensors 13 and their temporal changes by utilizing the ambiguity of the language. Such verbalization can be performed with an expression (nuance) close to the sense of the operator. In addition, the appearance can be verbalized and expressed also with respect to image data obtained from, for example, an image sensor.

  Moreover, since the measurement value of the sensor 13 vibrates by automatic control, such vibration can be well expressed by verbalization. Moreover, the operating state of the incinerator 11 can be represented as the verbalization state information 232a by comprehensively utilizing the time-series tendency of the measurement values of the various sensors 13. Then, the verbalization state information 232a and the control information 232b are associated with each other to construct the history linguisticization database 232. Here, as the verbalization state information 232a and the control information 232b, the operator's skill such as at what timing, at what timing, at what time, and at what rate, the operator changes the control value The control content related to can be recorded in the history linguisticization database 232.

  The current operation state of the incinerator 11 can be verbalized, and the past similar operation state can be inquired by referring to the history languageization database 232 based on the verbalized state information. Here, linguistic similarities can be exploited to reference similar past driving conditions. Therefore, even if there is no past driving state that matches the current verbalization state information, it is possible to propose control content candidates by applying similar ones with similar numbers. In addition, the history verbalization database 232 can be easily corrected later.

  As a result of referring to the history language conversion database 232, the control candidate identification unit 212 presents candidates for control content on the display unit 240. At this time, the reason for selecting the candidate can be displayed in accordance with the display unit 240. Thereby, the operator 21 can select an appropriate control content in consideration of the selection reason.

  As described above, the information processing apparatus 200A according to the present embodiment can select appropriate control content by grasping that the operation state of the facility is a precursor state of abnormality.

  On the other hand, for example, when referring to a database in which the past operation state is numerically stored based on the measurement values of the plurality of sensors 13 in the current operation state of the incinerator 11, the data amount of the database Can be huge and can take a great deal of time to reference. Therefore, an apparatus of such a configuration can not be put to practical use.

(Summary of other configurations and effects)
In the information processing apparatus 200A according to one aspect of the present invention, the control candidate identifying unit 212 stores a combination of type conceptualization information that matches or is similar to a combination of multiple types of type conceptualization information (type verbalization state information) at the current time It may be specified from the verbalization database 232), and control content candidates corresponding to the combination may be specified.

  According to the above configuration, it is possible to refer to the database based on a combination of a plurality of types of type conceptualization information, and more accurately inquire about the past operation state of the facility and the current operation state of the facility Can.

  In the information processing apparatus 200A according to one aspect of the present invention, a neural network processing unit (determination in which the control candidate identification unit 212 inputs a combination of the plurality of types of type conceptualization information (type linguistic state information) at the current time as multi-input information A section 212a) may be provided, and candidates for the control content corresponding to the combination may be specified based on the output result of the neural network processing section.

  In general, in the case where some judgment is performed based on a large number of input information, it is difficult to realize the judgment processing by an algorithm, and in many cases, the processing is very burdensome. That is, as described above, it is often difficult to construct an algorithm that performs processing by inputting a combination of multiple types of type conceptualization information as multi-input information.

  On the other hand, according to the above configuration, the neural network processing unit determines, based on the combination of the plurality of types of type conceptualization information, whether or not the operation state of the facility at the current time is a precursor state of abnormality. Ru. Once the neural network constructs a judgment processing unit by learning, subsequent judgment processing can be performed at high speed. In addition, the judgment processing unit can be constructed by learning without constructing a complicated algorithm. That is, according to the above configuration, it is possible to appropriately realize the processing unit that makes the above-described determination.

(Modification 1)
In another aspect of the present invention, the information processing apparatus is a control apparatus that controls the incinerator 11 as it is without presenting it to the operator based on the control content specified by the control candidate specification unit 212. Good. Thereby, the further automation of the operation of the refuse incineration facility 10 can be achieved.

(Modification 2)
The information processing apparatus 200A in the first embodiment of the present invention records verbalization control result information obtained by verbalizing the result of the manual intervention performed by the operator in the history linguisticization database 232. The result does not have to be verbalized. In a modification of the present invention, the time until the operation of the facility is normalized, or the degree to which the steady state is approached, etc., is used as the control result information as a result of the operator performing the manual intervention. The control result information is recorded in the history language conversion database 232, and is evaluated by the evaluation unit 212b.

Second Embodiment
Another embodiment of the present invention will be described based on FIGS. The configuration other than that described in the present embodiment is the same as that of the first embodiment. Further, for convenience of explanation, members having the same functions as the members shown in the drawing of the first embodiment above are given the same reference numerals, and the description thereof will be omitted.

  In the information processing apparatus 200A according to the first embodiment, the determination unit 212a performs neural network processing to determine whether the current operating condition of the incinerator 11 is a precursor condition of abnormality, and determines that it is a precursor condition. When it is determined, the control candidate specifying unit 212 specifies the control content candidate with reference to the history linguisticization database 232. On the other hand, the information processing apparatus 200B of the present embodiment is different in the following points. That is, the verbalization processing unit 211 integrates and verbalizes a plurality of types of state information related to the operating state of the incinerator 11 using the determination dictionary 234, and generates the verbalized state integrated information 232d. The determination unit 212c according to the present embodiment is configured to perform text matching (or text mining) on the history linguisticization database 232 based on the linguistic state integrated information 232d.

<Information processing device 200B>
An information processing apparatus 200B according to the present embodiment will be described with reference to FIG. FIG. 10 is a block diagram showing the main configuration of the information processing apparatus 200B of the present embodiment.

  As shown in FIG. 10, the information processing apparatus 200B includes the determination unit 212c in the control candidate identification unit 212, and includes the determination dictionary 234 in the storage unit 230. In addition, verbalized state integrated information 232 d is recorded in the history verbalization database 232. The other configuration in the information processing device 200B is the same as that of the information processing device 200A.

  The verbalization processing unit 211 in the present embodiment uses the verbalization rule data 231 to verbalize the operation state of the incinerator 11 based on various process data to generate verbalization state information. Then, based on the plurality of types of generated verbalization state information, using the determination dictionary 234, the verbalized state integration information (upper conceptualization information) 232d is generated by integrating the linguisticized state information.

(Language of driving condition)
The verbalization processing based on the received process data performed by the verbalization processing unit 211 described above will be described with reference to FIG. FIG. 11 is a diagram for describing generation of linguistic state integrated information 232 d based on process data.

  As shown in FIG. 11, the verbalization processing unit 211 first determines certain process data, for example, by categories such as "operating point level", "control margin", and "stability" Perform.

  For example, the operating point level can be verbalized as "very high", "high", "middle", "low". Also, for example, the control margin can be verbalized as “with margin”, “with”, “half”, “not much”, and the above stability is “very stable”, “stable” It can be verbalized like "normal" or "unstable".

  Then, the results of verbalization according to these categories are combined, for example, into sentences such as “It is at a higher place than the normal state and there is no margin for control and it is unstable”. Such textification is also performed on other various process data.

  Then, the verbalization processing unit 211 uses the determination dictionary 234 to comprehensively determine the operation state of the incinerator 11 based on the results of converting various process data into sentences, and as a conclusion, for example, “the apparent specific gravity is small It generates verbalized state integrated information 232 d that “the waste is oversupplied”.

  Although the specific aspect of the determination dictionary 234 is not particularly limited, for example, the determination dictionary 234 may be a list of combinations of results obtained by converting various process data into sentences and corresponding conclusions. Such a list can be created in advance.

  A plurality of such verbalized state integrated information 232 d may be generated. In general, when a person determines the operating state of the incinerator 11, several hypotheses may be made based on the information of various sensors 13. Similarly, even when the operating state of the incinerator 11 is comprehensively determined using the determination dictionary 234, it is natural that a plurality of types of verbalized state integrated information 232d are generated.

  That is, for example, from the operation state of the incinerator 11 at a certain point in time, plural types of verbalized state integrated information 232 d may be generated as follows. That is, the linguistic state integrated information 232 d can be generated based on the determination dictionary 234 using, for example, three types of process data among various process data. Then, other types of linguistic state integrated information 232 d may be generated based on the determination dictionary 234 using, for example, five other types of process data among various types of process data.

(Construction of history languageization database 232)
With reference to FIG. 12, the flow of the process of constructing the history linguisticization database 232 of the present embodiment using the existing control history data accumulated in the past operation of the refuse incineration facility 10 will be described. FIG. 12 is a flow chart diagram showing a flow of processing for constructing the history languageization database 232 using existing control history data.

  As shown in FIG. 12, first, the existing control history data accumulated is prepared (S41). The information processing device 200B reads this control history data.

  The control unit 210 extracts information on one intervention log from the information on the plurality of intervention logs included in the control history data (S42).

  The verbalization processing unit 211 performs verbalization processing on the extracted intervention log as follows. That is, using the verbalization rule data 231, the operating state is verbalized based on various process data of the waste incineration facility 10 from the point of time when control was performed to the first predetermined time before the time to make verbalized state information (S43). The verbalization state information is generated for each process data and is a plurality of types of type conceptualization information. The verbalization processing unit 211 generates verbalized state integrated information 232 d using the determination dictionary based on the plurality of types of type conceptualization information (S 44). A plurality of types of linguistic state integrated information 232 d may be generated.

  Next, the verbalization processing unit 211 verbalizes the operation state into verbalization control result information based on various process data of the waste incineration facility 10 from the time when control is performed in the past to the second predetermined time later. (S45).

  When a plurality of types of linguistic state integrated information 232 d are generated, the recording processing unit 220 is the most improved verbalized out of the plurality of types of verbalized state integrated information 232 d based on the linguistic control result information. The state integration information 232 d is extracted. That is, the linguistic state integrated information 232 d most improved, the control content at that point in time, and the verbalization control result information are associated with each other and recorded in the history verbalization database 232 (S 46).

  Here, the following method (i) or (ii) may be used as a method of extracting the most improved verbalized state integrated information 232d.

  (I) A certain verbalized state integrated information 232 d and, for example, the sensor 13 corresponding to the verbalized state integrated information 232 d are defined in advance. Based on the manual control performed by the operator 121, it is detected which measurement value of the sensor 13 has fluctuated, and the most improved (close to the stable state) measurement value is extracted. The verbalized state integrated information 232d corresponding to the sensor 13 indicating the extracted measurement value can be extracted as the most improved verbalized state integrated information 232d.

  (Ii) The recording processing unit 220 presents, for example, on the display unit 240, a plurality of types of linguistic state integrated information 232d, and control result information or verbalization control result information as numerical data. Then, for example, the skilled operator 21 can look at the result information and select the linguistically integrated state information 232 d expected to be most improved.

  The processes in S42 to S45 can be performed for each of a plurality of intervention logs included in the control history data to build up the history verbalization database 232.

<Operation Selection Process Performed by Information Processing Device 200B>
A flow of information processing by the information processing device 200B will be described with reference to FIG. FIG. 13 is a flowchart showing the flow of information processing by the information processing apparatus 200B.

  As shown in FIG. 13, first, in the normal state, the same processes as S11 and S12 (see FIG. 8) are performed as S51 and S52, respectively.

  Next, when the predetermined time has passed (YES in S52), the verbalization processing unit 211 verbalizes the current operating state of the incinerator 11 using various process data, and generates verbalization state information (S53).

  Furthermore, the verbalization processing unit 211 integrates various types of verbalization state information using the determination dictionary 234, and generates verbalized state integration information 232d (S54).

  The determination unit 212a refers to the history linguisticization database 232 based on the verbalization state integrated information 232d generated by the verbalization processing unit 211, and performs text matching or text mining (S55). That is, it is inquired whether or not the verbalized state integrated information 232 d recorded in the history verbalization database 232 corresponds to the verbalized state integrated information 232 d generated in S 54.

  If there is no corresponding linguistic state integrated information 232 d (NO in S56), the processing from S51 is repeated again. When there is the corresponding linguistic state integrated information 232 d (YES in S 56), the control candidate specifying unit 212 displays the candidates of the control content on the display unit 240 in the order of evaluation based on the corresponding linguistic state integrated information (S 57) ). The corresponding linguistic state integrated information may be identical or similar.

  Thereafter, as S58 to S60, processes similar to the above S18 to S20 (see FIG. 8) are performed.

(Summary of other configurations and effects)
In the information processing apparatus 200 </ b> B in one aspect of the present invention, the conceptualization processing unit (language conversion processing unit 211) combines a plurality of types of type conceptualization information (type language conversion status information) as the conceptualization information (language conversion status information). One superordinate conceptualization information (verbalization state integrated information 232d) may be generated.

  According to the above configuration, the information recorded in the database can be superordinate conceptualization information in which a plurality of types of type conceptualization information are put together. Therefore, the control candidate identification unit can easily refer to the database.

  In the information processing apparatus 200B according to one aspect of the present invention, upper conceptualization information (verbalization state integration information) in which the conceptualization processing unit (verbalization processing unit 211) combines the plurality of types of type conceptualization information (classification linguistic state information). Two or more types of the candidate 232 d) may be generated.

  According to the above configuration, the state of the facility can be expressed by a plurality of types of the upper-level conceptualization information. Here, when a person determines the operating state of the equipment, there are cases in which several equipment states are assumed. The state of equipment can be expressed in such a form close to human judgment.

  In the information processing apparatus 200B according to one aspect of the present invention, the recording processing unit 220 determines that the upper-level candidate information (verbalization state integration information 232d) generated by the conceptualization processing unit (verbalization processing unit 211) is higher What improved the result in which the control information corresponding to conceptualization information was performed may be matched with this control information, and you may record on a database (history linguisticization database 232).

  According to the above configuration, the higher-level conceptualization information recorded in the database and the control content associated therewith are highly likely to improve the result by performing the control content. Therefore, the control content specified by the control candidate specifying unit can be made more appropriate.

  In the information processing apparatus 200B according to one aspect of the present invention, the control candidate identification unit 212 is a database (a history linguistic database) that is higher-level conceptualization information that matches or is similar to the higher-level conceptualization information (language integration state integration information 232d) at the current time 232), and the control content candidate corresponding to the superordinate conceptualization information may be specified.

  According to the above configuration, the control content candidate specified by the control candidate specification unit has a high probability that the result will be improved by performing the control content. Therefore, it is possible to realize an information processing apparatus capable of selecting more appropriate control content in a precursor state of an abnormality in the operation of the facility.

(Modification)
In the information processing apparatus 200B according to the second embodiment, when constructing the history linguisticization database 232, the recording processing unit 220 selects the most improved linguistically integrated state information 232d out of a plurality of types of linguistically integrated state information 232d. It was extracted. On the other hand, in another aspect of the present invention, the most improved verbalized state integrated information 232d is selected as follows. That is, on the display unit 120b of the operation control device 120, a plurality of types of verbalized state integrated information 232d as a result of determining the operating state of the incinerator 11 at that time are sequentially displayed. Then, when the operator 121 performs the manual intervention, in parallel with it, the verbalized state integrated information 232 d which the operator 121 judges to be most appropriate is selected. As a result, it is possible to obtain a combination of data in which the verbalized state integrated information 232 d determined to be most appropriate and the control content and the control result are associated. By storing this combination, the history languageization database 232 may be constructed.

[Example of software implementation]
The control blocks (in particular, the control unit 210) of the information processing apparatus 200A and the information processing apparatus 200B may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like. It may be realized by software using.

  In the latter case, the information processing apparatus 200A and the information processing apparatus 200B are a CPU that executes instructions of a program that is software that implements each function, and a ROM in which the program and various data are readably recorded by a computer (or CPU). (Read Only Memory) or a storage device (these are referred to as "recording media"), a RAM (Random Access Memory) for developing the program, and the like. The object of the present invention is achieved by the computer (or CPU) reading the program from the recording medium and executing the program. As the recording medium, a “non-transitory tangible medium”, for example, a tape, a disk, a card, a semiconductor memory, a programmable logic circuit or the like can be used. The program may be supplied to the computer via any transmission medium (communication network, broadcast wave, etc.) capable of transmitting the program. The present invention can also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission.

  The present invention is not limited to the above-described embodiments, and various modifications can be made within the scope of the claims, and embodiments obtained by appropriately combining the technical means disclosed in the different embodiments. Is also included in the technical scope of the present invention.

  10: waste incineration facility (equipment) 11: incinerator (equipment) 200A, 200B: information processing apparatus 210: control unit (control processing unit) 211: verbalization processing unit (conceptification processing unit) 212: control candidate identification unit 212a: Judgment unit (neural network processing unit) 212b: evaluation unit 220: recording processing unit 232b: control information 232: history linguisticization database (database) 250: input unit (selection unit)

Claims (9)

In an information processing apparatus in a facility that performs process control,
A conceptualization processing unit that generates conceptualization information that conceptualizes state information related to the operating state of the facility;
A recording processing unit that records control information related to a predetermined type of control content for the facility and the conceptualization information related to the control information in a database in association with each other;
An information processing apparatus comprising: a control candidate specifying unit for specifying a candidate for control content to be performed on the facility by referring to the database based on the conceptualization information generated by the conceptualization processing unit .   The information processing apparatus according to claim 1, wherein the conceptualization processing unit generates linguistic data representing the state information in a language as the conceptualization information.   The information processing according to claim 1 or 2, further comprising: an evaluation unit that specifies a plurality of candidates for the control content and ranks the candidates based on a predetermined criterion. apparatus.   The said conceptualization process part produces | generates the type conceptualization information about each of several types of status information which shows the driving | running state of the said installation as said conceptualization information, The claim is characterized by the above-mentioned. Information processing equipment. The information processing apparatus according to any one of claims 1 to 4.
A selection unit that allows the user to select the control content;
And a control processing unit that executes the selected control content. The information processing apparatus according to any one of claims 1 to 4.
A control processing unit configured to execute the control content specified by the control candidate specifying unit; In a control method of an information processing apparatus in a facility performing process control,
A conceptualization processing step of generating conceptualization information that conceptualizes state information regarding the operation state of the equipment;
A recording processing step of recording control information related to a predetermined type of control content for the facility and the conceptualization information related to the control information in a database in association with each other;
A control candidate specifying step of specifying a candidate for control content to be performed on the facility by referring to the database based on the generated conceptualization information during operation of the facility. Device control method.   A control program for causing a computer to function as the information processing apparatus according to any one of claims 1 to 4, wherein the control program for causing a computer to function as each of the units.   The computer-readable recording medium which recorded the control program of Claim 8.

Images