JP2018163542A - Prediction device, prediction system, prediction method, and prediction program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily predict a transition associated with a time change of a state of a physical system such as a plant.SOLUTION: A prediction device comprises: an operation information storage unit that stores operation information including at least an element value of a physical system; an information collecting unit that collects operation information of the physical system and stores the operation information in the operation information storage unit; an element value prediction unit that calculates a future element value of the physical system as a predicted element value on the basis of the element value; and an operation state prediction unit that predicts a future operation state of the physical system on the basis of the predicted element value.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a prediction device, a prediction system, a prediction method, and a prediction program.

  A technique for predicting the future state of a plant based on current operation information of a physical system such as a plant (hereinafter also simply referred to as a plant) is known. Such techniques typically predict future conditions by generating a predictive model based on measured values of specific parameters of the plant.

  Patent Document 1 discloses an apparatus for accurately estimating / predicting a plant state using a mathematical model for various plant states. Specifically, in Patent Document 1, in order to estimate the state of a plant, the user himself / herself inputs a plant data collection period, a prediction model parameter, a relevance calculation parameter, a plant state quantity selection threshold, and the like. ing.

  Japanese Patent Application Laid-Open No. 2004-228561 discloses a technique that can perform detailed diagnosis of a prediction result and that selects a model with high accuracy while adapting to changes in an information source. Specifically, each time patent data 2 receives time series data, it generates a model series for predicting the occurrence of each data, and among the generated model series, a model series having the smallest error from an actual measurement value. Is selected.

  Patent Document 3 supports the process model parameter adjustment by visualizing the change of the process model parameter and executing the simulation so that the effect of the change on the entire plant can be easily grasped. An apparatus is disclosed. Specifically, Patent Document 3 is configured so that the amount of change in the relevance ratio in both positive and negative directions over the entire plant when any model parameter is changed can be grasped.

JP 2000-181526 A JP 2008-003920 A JP 2010-146137 A

  As described above, Patent Documents 1 to 3 generate optimal prediction models by arbitrarily changing parameters in the plant in order to predict the state of the plant. In this case, as the number of parameters that affect the state of the plant increases, it may be difficult to select parameters to be changed to generate a model.

  An object of the present invention is to provide a prediction device, a prediction system, a prediction method, and a prediction program capable of easily predicting a transition accompanying a time change of a state of a physical system such as a plant.

  The prediction apparatus according to the first aspect of the present invention includes an operation information storage unit that stores operation information including at least an element value of a physical system, and collects the operation information of the physical system and uses the operation information as the operation information. An information collection unit stored in a storage unit; an element value prediction unit that calculates the future element value of the physical system as a predicted element value based on the element value; and a future of the physical system based on the predicted element value An operating state prediction unit for predicting the operating state of

  A prediction system according to a second aspect of the present invention includes the prediction apparatus according to the first aspect of the present invention and a monitoring device that monitors a physical system, and the monitoring device includes at least an element value of the physical system. A measurement unit for measuring information is provided.

  The prediction method according to the third aspect of the present invention holds operation information including at least element values of a physical system, collects the operation information of the physical system, and stores the operation information in an operation information storage unit. The future element value of the physical system is calculated as a predicted element value based on the element value, and the future operating state of the physical system is predicted based on the predicted element value.

  The prediction program according to the fourth aspect of the present invention includes a process for storing operation information including at least element values of a physical system in a computer, and storing the operation information while collecting the operation information of the physical system. A process of storing in the storage unit, a process of calculating the future element value of the physical system as a predicted element value based on the element value, and a future operating state of the physical system is predicted based on the predicted element value Process.

  According to the present invention, it is possible to provide a prediction device, a prediction system, a prediction method, and a prediction program capable of easily predicting a transition accompanying a time change of a state of a physical system such as a plant.

It is a block diagram which shows the structure of the prediction apparatus which concerns on the 1st Embodiment of this invention. It is a flowchart which shows the flow of operation | movement of the prediction apparatus which concerns on the 1st Embodiment of this invention. It is a block diagram which shows the structure of the prediction system which concerns on the 2nd Embodiment of this invention. It is a flowchart which shows the flow of operation | movement until the prediction system which concerns on the 2nd Embodiment of this invention produces | generates a correlation model. It is a figure which shows an example of the operation information of a physical system. It is a figure which shows an example of a correlation model. It is a block diagram which shows the structure of the prediction system which concerns on the related technology of the 2nd Embodiment of this invention. It is a flowchart which shows the flow of operation | movement of the prediction system which concerns on the related technology of the 2nd Embodiment of this invention. It is the schematic diagram which showed the connection between the elements in a plant, (a) has shown the connection between the elements in a normal state, (b) has shown the mode of destruction of a correlation. It is a block diagram which shows the structure of the prediction system which concerns on the other related technique of the 2nd Embodiment of this invention. It is a flowchart which shows the flow of operation | movement until the prediction system which concerns on the 2nd Embodiment of this invention calculates the predicted value of an element. It is a flowchart which shows the flow of operation | movement until the prediction system which concerns on the 2nd Embodiment of this invention estimates the operation information of a plant. It is a block diagram which shows the structure of the prediction apparatus which concerns on other embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals, and description thereof will be omitted as appropriate.

[First Embodiment]
FIG. 1 is a block diagram showing the configuration of the prediction apparatus according to the first embodiment of the present invention.

  The prediction device 10 includes an operation information storage unit 101, an information collection unit 102, an element value prediction unit 103, and an operation state prediction unit 104.

  The operation information storage unit 101 holds operation information including at least element values of the plant. In the present embodiment, the “element” means a measuring instrument such as a flow rate sensor, a pressure sensor, and a current sensor installed in the plant. The “element value” means a measured value of a flow sensor, a pressure sensor, a current sensor, or the like. The operation information storage unit 101 may hold all operation information from the past to the present, or may hold operation information only for a specific fixed period.

  The information collection unit 102 collects plant operation information. The information collecting unit 102 stores the collected operation information in the operation information storage unit 101.

  The element value prediction unit 103 calculates a predicted element value, which is a future element value of the plant, based on the element value in the operation information held by the operation information storage unit 101.

  The operation state prediction unit 104 predicts the future operation state of the plant based on the prediction element value.

  FIG. 2 is a flowchart showing an operation flow of the prediction apparatus 10. Hereinafter, the operation flow of the prediction device 10 will be described with reference to FIGS. 1 and 2.

  First, the operation information storage unit 101 holds plant operation information (step S101). Next, the information collection unit 102 collects plant operation information and stores it in the operation information storage unit 101 (step S102). Next, the element value prediction unit 103 calculates a prediction element value based on the element value of the operation information (step S103). Finally, the operation state prediction unit 104 predicts the future operation state of the plant based on the prediction element value (step S104).

[Second Embodiment]
FIG. 3 is a block diagram showing a configuration of a prediction system according to the second embodiment of the present invention.

  The prediction system 100A includes a prediction device 10A, a monitoring device 20, and an administrator device 30. A monitoring device 20 and an administrator device 30 are connected to the prediction device 10A.

  The prediction apparatus 10A includes an operation information storage unit 101, an information collection unit 102, an element value prediction unit 103, an operation state prediction unit 104, a prediction element value storage unit 105, a correlation model generation unit 106, and a correlation model storage. Unit 107 and instruction unit 108.

  The information collection unit 102 receives operation information regarding the plant from the monitoring device 20. The information collecting unit 102 stores the received operation information in the operation information storage unit 101.

  The element value prediction unit 103 extracts element values held by the operation information storage unit 101. Further, the element value prediction unit 103 calculates a predicted element value based on the extracted element value. Furthermore, the element value prediction unit 103 stores the calculated prediction element value in the prediction element value storage unit 105. The element extracted by the element value prediction unit 103 is not particularly limited, and the user may select an element that is arbitrarily extracted. Specifically, the element extracted by the element value prediction unit 103 is, for example, the operating state of a turbine or a boiler included in the plant.

  The prediction element value storage unit 105 holds a prediction element value.

  The correlation model generation unit 106 extracts operation information for a certain period from the operation information storage unit 101. Then, the correlation model generation unit 106 derives a time-series conversion function for at least two arbitrary values of operating information among the extracted operating information. Thereby, the correlation model production | generation part 106 can produce | generate the correlation model of the operating state of a plant.

  The correlation model storage unit 107 holds the correlation model generated by the correlation model generation unit 106. That is, the correlation model generation unit 106 stores the generated correlation model in the correlation model storage unit 107.

  The instruction unit 108 transmits an instruction from the user received by the administrator device 30 to the monitoring device 20.

  The operation state prediction unit 104 predicts the future operation state of the plant based on the prediction element value read from the prediction element value storage unit 105 and the correlation model read from the correlation model storage unit 107. Specifically, the operating state prediction unit 104 predicts the future operating state of the plant by receiving a specific prediction element value for the correlation model held by the correlation model storage unit 107.

  The monitoring device 20 is connected to the plant and includes a measurement unit 201 that measures the operation state of the plant such as the flow rate, temperature, and pressure of the plant. The measuring unit 201 and the monitoring device 20 execute process control of the plant in accordance with an instruction from the instruction unit 108 of the prediction apparatus 10A. Note that the monitoring device 20 may be an information processing device that provides the operating state of the plant to the user through a web service or a business service.

  The administrator device 30 receives the prediction result of the operating state prediction unit 104 of the prediction device 10A and displays the prediction result to the user. Further, the administrator device 30 receives an instruction from the user according to the prediction result, and transmits the received instruction to the instruction unit 108 of the prediction device 10A.

  FIG. 4 is a flowchart showing a processing flow until the prediction system 100A generates a correlation model. Hereinafter, the flow of processing until the prediction system 100A generates a correlation model will be described with reference to FIGS. 3 and 4.

  First, the information collection unit 102 collects operation information indicating the operation state of the plant from the monitoring device 20 and stores it in the operation information storage unit 101 (step S201). Here, when the monitoring device 20 is executing process control on the plant, the information collection unit 102 collects operation information at regular intervals and stores it in the operation information storage unit 101 in time series. can do. That is, the information collecting unit 102 can continuously collect plant operation information.

  FIG. 5 is a diagram illustrating an example of plant operation information held by the operation information storage unit 101. In FIG. 5, “Time” indicates the date and time when the operating state is collected, “A.OUT_FLOW” indicates the flow rate at the outlet of the pump A, “B.OUT_FLOW” indicates the flow rate at the outlet of the pump B, and “Z. “OUT_FLOW” indicates the flow rate at the outlet of the pump Z. Specifically, the flow rate of the pump A at 0:00 on August 1, 2016 is 12. FIG. 5 shows that the flow rates of the respective pumps are collected at intervals of 1 minute. For example, the flow rate of pump A at 0:01 on August 1, 2016 is 15, the flow rate of pump A at 0:02 on August 1, 2016 is 34, and the flow rate of pump A on August 1, 2016 is 0:03. The flow rate of pump A in minutes is 63. In addition, in FIG. 5, although the operation information for every 1 minute interval is shown, this is an illustration and does not limit the present invention. The interval at which the operation information is collected may be arbitrary.

  Refer to FIG. 4 again. Next, the correlation model generation unit 106 generates a correlation model by deriving a conversion function between pieces of operation information held by the operation information storage unit 101 (step S202).

  FIG. 6 is a diagram illustrating an example of a correlation model generated by the correlation model generation unit 106. In FIG. 6, the conversion function between the input (X) and the output (Y) can be expressed by, for example, the following expression (1).

  In FIG. 6, α in equation (1) when the input (X) is “A.OUT_FLOW” and the output (Y) is “B.OUT_PRESSURE” is “−0.6”, and β is “100”. It is. Further, the correlation model generation unit 106 calculates the weight based on the conversion error that is the difference between the time series of the output (Y) value converted by Expression (1) and the time series of the actually measured value of the output (Y). can do. Specifically, when the input (X) is “A.OUT_FLOW” and the output (Y) is “B.OUT_PRESSURE”, the weight is “0.88”. Similarly, the correlation model generation unit 106 derives a conversion function between any two pieces of operation information, and extracts a conversion function having a predetermined weight as an effective correlation. Thereby, the correlation model production | generation part 106 can produce | generate a correlation model as shown in FIG. The above formula (1) is an exemplification, and does not limit the conversion function in the present embodiment. In the present embodiment, the conversion function is not particularly limited as long as it converts a time series of values of any two pieces of operation information.

  Refer to FIG. 4 again. Finally, the correlation model generation unit 106 stores the generated correlation model in the correlation model storage unit 107 (step S203).

  Here, before explaining this embodiment in more detail, the related technique of this embodiment is demonstrated. FIG. 7 is a block diagram illustrating a configuration of a prediction system 100B according to the related technology of the present embodiment. The prediction system 100B includes a prediction device 10B, a monitoring device 20, and an administrator device 30. The prediction device 10B is a prediction device that analyzes a plant failure using a correlation model.

  The prediction device 10B includes an operation information storage unit 101, an information collection unit 102, a correlation model generation unit 106, a correlation model storage unit 107, an instruction unit 108, a correlation change analysis unit 109, and a failure analysis unit 110. Prepare.

  FIG. 8 is a flowchart showing the flow of operations of the correlation change analysis unit 109 and the failure analysis unit 110 after the correlation model generated by the correlation model generation unit 106 is stored in the correlation model storage unit 107 in the prediction device 10B. It is. Hereinafter, with reference to FIG. 7 and FIG. 8, the flow of operations of the correlation change analysis unit 109 and the failure analysis unit 110 will be described.

  The correlation change analysis unit 109 calculates the destruction distribution of the correlation model in the normal state of the plant after the correlation model generation unit 106 generates the correlation model. Here, in the correlation between a plurality of parameters in the normal operating state of the plant, the fact that the correlation cannot be maintained is referred to as destruction of the relationship. A collection of destruction information of this relationship is called a destruction distribution.

  FIG. 9 is a schematic diagram showing the connection between elements in the plant, FIG. 9 (a) shows the connection between elements in a normal state, and FIG. 9 (b) shows how the correlation is broken. Yes. FIG. 9A shows that when the plant is in a normal state, element A, element B, element C, element D, element E, and element F are connected, and there is a correlation between the elements. Is shown. In FIG. 9B, the element D, the element E, and the element F are connected by a broken line, which means that the relationship between the element D, the element E, and the element F is broken. The correlation change analysis unit 109 calculates a fracture distribution based on information regarding the destruction of the relationship between elements as shown in FIG.

  After the correlation model generation unit 106 generates the correlation model, the correlation change analysis unit 109 calculates a deviation between the operation information newly collected by the information collection unit 102 and the correlation indicated in the correlation model (step S301). . For example, in FIG. 5, it is assumed that data of “August 1, 2016 23:59” is new operation information. In this case, the correlation change analysis unit 109 calculates the value of the operation information at “23:59 on August 1, 2016” calculated using the conversion function in the correlation model shown in FIG. 6 and “August 1, 2016”. The difference (deviation) from the measured value of the operation information at “23:59” is calculated.

  Next, the correlation change analysis unit 109 calculates whether or not the difference between the value of the operation information calculated using the conversion function and the actual value of the operation information exceeds a predetermined threshold (step S302). Here, if the difference is smaller than the threshold (“NO” in step S302), correlation change analysis section 109 determines that the correlation is maintained, and returns to step S301. Further, when the difference is larger than the threshold value (“YES” in step S302), correlation change analysis section 109 determines that correlation destruction has occurred, abnormal degree information indicating the degree of correlation change, and correlation change Correlation change information including abnormal element information indicating an element related to the information is created (step S303). In this case, the correlation change analysis unit 109 outputs the correlation change information to the failure analysis unit 110. The correlation change analysis unit 109 performs the above-described process on all newly acquired operation information, and determines whether there is a correlation change.

  If the degree of change exceeds a predetermined threshold based on the correlation change information received from the correlation change analysis unit 109, the failure analysis unit 110 sends a plant to the user via the administrator device 30. Is presented that there is a possibility that an abnormality has occurred (step S304).

  FIG. 10 is a block diagram illustrating a configuration of a prediction system 100C according to another related technology. The prediction system 100C includes a prediction device 10C, a monitoring device 20, and an administrator device 30.

  The prediction device 10 </ b> C includes an operation information storage unit 101, an information collection unit 102, an operation state prediction unit 104, a correlation model generation unit 106, and a correlation model storage unit 107. In such a prediction apparatus 10 </ b> C, the operating state prediction unit 104 first receives a correlation model held by the correlation model storage unit 107. Moreover, the operation state prediction unit 104 simulates operation information of the entire plant for one time of one element. Then, the operating state prediction unit 104 outputs the simulation result to the administrator device 30.

  As described above, the prediction device 10B and the prediction device 10C have one time that can be predicted. However, for planning a plant maintenance plan, it is preferable to calculate information regarding the continuous transition of the operating state of the plant.

  FIG. 11 is a flowchart showing a flow of operations until the element value prediction unit 103 calculates an element value after the correlation model generated by the correlation model generation unit 106 is stored in the correlation model storage unit 107 in the prediction system 100A. is there. Hereinafter, the operation flow of the prediction system 100A according to the present embodiment will be described with reference to FIG. 3 and FIG.

  After the correlation model generation unit 106 generates a correlation model, the element value prediction unit 103 extracts time series data of operation information for a specific period of a specific element from the operation information storage unit 101 (step S401). . Time series data is often non-stationary due to various factors such as seasonality and noise. Therefore, it is preferable to make time series data stationary.

  Next, the element value prediction unit 103 determines whether or not the variance of the time series data is constant (step S402). If the variance is not constant (“NO” in step S402), the element value prediction unit 103 performs exponential transformation or logarithmic transformation on the time series data to make the variance constant (step S403).

  Next, the element value prediction unit 103 determines whether or not the time series data is stationary data (step S404). If the time series data is non-stationary (“NO” in step S404), non-stationarity is removed (step S405).

  Next, the element value prediction unit 103 determines whether the time series data has seasonality (periodicity) (step S406). If the time series data has seasonality (“YES” in step S406), the seasonality is removed (step S407).

  Stationary data can be created by the processing in steps S401 to S407. However, there is a possibility that the element value prediction unit 103 may make an incorrect prediction because the chance is not eliminated if only one period is calculated. Therefore, after “YES” in step S402, “YES” in step S404, “NO” in step S406, and step S407, the element value prediction unit 103 creates steady data for a plurality of periods (step S408).

  Next, the element value prediction unit 103 performs steady data according to an AR model (Autoregressive Model) or an ARMA model (Autoregressive moving average model) with respect to stationary data for a plurality of periods. Then, the contingency is excluded and the predicted element value is calculated (step S409). In general, the AR model is described in the form of the following expression (2), and the ARMA model is described in the form of the following expression (3).

In Expression (2), Y is a predicted value of an element, and X _{1 to} X _n are past values of the element value. A _{1 to An} are coefficients of X _{1 to} X _n , respectively. Moreover, Formula (2) may include a constant term.

  In Expression (3), Y is a predicted value of an element, and X1 to Xn are past values of the element value. A1 to An are coefficients of X1 to Xn, respectively. ε1 to εn are error terms, respectively. B1 to Bn are coefficients of ε1 to εn, respectively. Moreover, Formula (3) may include a constant term.

  FIG. 12 is a flowchart showing a flow from when the element value prediction unit 103 calculates a prediction element value until the operating state prediction unit predicts the future state of the plant.

  First, the element value prediction unit 103 calculates a prediction element value, and then stores the calculated prediction element value in the prediction element value storage unit 105 (step S501).

  Next, the operating state prediction unit 104 receives a prediction element value at one time from the prediction element value storage unit 105, and receives a correlation model from the correlation model storage unit 107 (step S502).

  Next, the operating state prediction unit 104 calculates other element values so as to maintain the correlation shown in the correlation model (step S503).

  Finally, the operating state predicting unit 104 repeats the processes of Step S502 and Step S503, and calculates other element values continuously, thereby calculating a temporal change in the operating state of the plant (Step S504).

  As described above, according to the present embodiment, it is easy to simulate a time change of a plant based on element values of a plurality of periods. Therefore, this embodiment can confirm how the operating state of the plant will change in the future, and will help to devise a maintenance plan so that maintenance is executed at an optimal time. .

<Other embodiments>
The prediction device may be realized by hardware or software. The prediction device may be realized by a combination of hardware and software.

  FIG. 13 is a block diagram illustrating an example of an information processing device (computer) that configures the prediction device.

  As illustrated in FIG. 13, the information processing device 200 includes a control unit (CPU: Central Processing Unit) 210, a storage unit 220, a ROM (Read Only Memory) 230, a RAM (Random Access Memory) 240, and a communication interface. 250 and a user interface 260.

  The control unit (CPU) 210 can realize various functions of the prediction device by developing a program stored in the storage unit 220 or the ROM 230 in the RAM 240 and executing the program. The control unit (CPU) 210 may include an internal buffer that can temporarily store data and the like.

  The storage unit 220 is a large-capacity storage medium that can hold various data, and can be realized by a storage medium such as an HDD (Hard Disk Drive) and an SSD (Solid State Drive). The storage unit 220 may be a cloud storage that exists on the communication network when the information processing apparatus 200 is connected to the communication network via the communication interface 250. The storage unit 220 may hold a program that can be read by the control unit (CPU) 210.

  The ROM 230 is a non-volatile storage device that can be configured with a flash memory or the like having a smaller capacity than the storage unit 220. The ROM 230 may hold a program that can be read by the control unit (CPU) 210. Note that a program readable by the control unit (CPU) 210 may be held by at least one of the storage unit 220 and the ROM 230.

  Note that the program readable by the control unit (CPU) 210 may be supplied to the information processing apparatus 200 in a state of being temporarily stored in various computer-readable storage media. Such a storage medium is, for example, a magnetic tape, a magnetic disk, a magneto-optical disk, a CD-ROM, a CD-R, a CD-R / W, or a semiconductor memory.

  The RAM 240 is a semiconductor memory such as a DRAM (Dynamic Random Access Memory) and an SRAM (Static Random Access Memory), and can be used as an internal buffer for temporarily storing data and the like.

  The communication interface 250 is an interface that connects the information processing apparatus 200 and a communication network via a wired or wireless connection.

  The user interface 260 is, for example, a display unit such as a display and an input unit such as a keyboard, a mouse, and a touch panel.

  A part or all of the above embodiments can be described as in the following supplementary notes, but is not limited thereto.

[Appendix 1]
An operation information storage unit that stores operation information including at least element values of the physical system;
An information collection unit that stores the operation information in the operation information storage unit while collecting the operation information of the physical system;
An element value prediction unit that calculates the future element value of the physical system as a prediction element value based on the element value;
An operation state prediction unit that predicts a future operation state of the physical system based on the prediction element value.

[Appendix 2]
The prediction apparatus according to appendix 1, wherein the element value prediction unit calculates the prediction element value according to an autoregressive model.

[Appendix 3]
A prediction element value storage unit for holding the prediction element value;
The prediction device according to appendix 1 or 2, wherein the element value prediction unit stores the calculated prediction element value in the prediction element value storage unit.

[Appendix 4]
A correlation model generation unit configured to generate a correlation model indicating a correlation between the past operation information and the current operation information based on a time change of the past operation information held by the operation information storage unit; ,
The said operation state prediction part is a prediction apparatus as described in any one of appendix 1-3 which estimates the future operation state of the said physical system based on the said prediction element value and the said correlation model.

[Appendix 5]
The prediction device according to appendix 4, wherein the element value prediction unit calculates the prediction element value so as to maintain the correlation.

[Appendix 6]
Including the prediction device according to any one of appendices 1 to 5, and a monitoring device that monitors the physical system;
The monitoring device
A prediction system comprising a monitoring unit that monitors operation information including at least element values of the physical system.

[Appendix 7]
Holds operational information including at least element values of the physical system,
While collecting the operation information of the physical system, store the operation information in the operation information storage unit,
Calculating the future element value of the physical system based on the element value as a predicted element value;
A prediction method for predicting a future operating state of the physical system based on the prediction element value.

[Appendix 8]
The prediction method according to appendix 7, wherein the prediction element value is calculated according to an autoregressive model.

[Appendix 9]
Holding the predicted element value;
The prediction method according to appendix 7 or 8, wherein the calculated prediction element value is stored in a prediction element value storage unit.

[Appendix 10]
Based on the time change of the past operation information held by the operation information storage unit, generate a correlation model indicating a correlation between the past operation information and the current operation information,
The prediction method according to any one of appendices 7 to 9, wherein a future operating state of the physical system is predicted based on the prediction element value and the correlation model.

[Appendix 11]
The prediction method according to appendix 10, wherein the prediction element value is calculated so as to maintain the correlation.

[Appendix 12]
On the computer,
A process for holding operation information including at least element values of the physical system;
A process of storing the operation information in the operation information storage unit while collecting the operation information of the physical system;
A process of calculating the future element value of the physical system as a predicted element value based on the element value;
The prediction program which performs the process which estimates the future operating state of the said physical system based on the said prediction element value.

[Appendix 13]
In addition to the computer,
13. The prediction program according to appendix 12, which causes a process of calculating the prediction element value according to an autoregressive model.

[Appendix 14]
In addition to the computer,
Processing for holding the predicted element value;
The prediction program according to appendix 12 or 13, which causes the calculated prediction element value to be stored in a prediction element value storage unit.

[Appendix 15]
In addition to the computer,
A process of generating a correlation model indicating a correlation between the past operation information and the current operation information, based on a time change of the past operation information held by the operation information storage unit;
The prediction program according to any one of appendices 12 to 14, wherein a process for predicting a future operating state of the physical system is executed based on the prediction element value and the correlation model.

[Appendix 16]
In addition to the computer,
The pre-program according to appendix 15, wherein a process for calculating the prediction element value so as to maintain the correlation is executed.

10, 10A, 10B, 10C ... Prediction device 20 ... Monitoring device 30 ... Administrator device 100A, 100B, 100C ... Prediction system 101 ... Operating information storage unit 102 ... Information collection unit DESCRIPTION OF SYMBOLS 103 ... Element value prediction part 104 ... Operation | movement state prediction part 105 ... Prediction element value memory | storage part 106 ... Correlation model production | generation part 107 ... Correlation model memory | storage part 108 ... Instruction | indication part 109 ...・ Correlation change analysis unit 110 ・ ・ ・ Failure analysis unit

Claims (10)

An operation information storage unit that stores operation information including at least element values of the physical system;
An information collection unit that stores the operation information in the operation information storage unit while collecting the operation information of the physical system;
An element value prediction unit that calculates the future element value of the physical system as a prediction element value based on the element value;
An operation state prediction unit that predicts a future operation state of the physical system based on the prediction element value.   The prediction device according to claim 1, wherein the element value prediction unit calculates the prediction element value according to an autoregressive model. A prediction element value storage unit for holding the prediction element value;
The prediction device according to claim 1, wherein the element value prediction unit stores the calculated prediction element value in the prediction element value storage unit. A correlation model generation unit configured to generate a correlation model indicating a correlation between the past operation information and the current operation information based on a time change of the past operation information held by the operation information storage unit; ,
The prediction device according to claim 1, wherein the operation state prediction unit predicts a future operation state of the physical system based on the prediction element value and the correlation model.   The prediction device according to claim 4, wherein the element value prediction unit calculates the prediction element value so as to maintain the correlation. Including the prediction device according to any one of claims 1 to 5 and a monitoring device that monitors a physical system,
The monitoring device
A prediction system comprising a measurement unit that measures operation information including at least element values of the physical system. Holds operational information including at least element values of the physical system,
While collecting the operation information of the physical system, store the operation information in the operation information storage unit,
Calculating the future element value of the physical system based on the element value as a predicted element value;
A prediction method for predicting a future operating state of the physical system based on the prediction element value.   The prediction method according to claim 7, wherein the prediction element value is calculated according to an autoregressive model. On the computer,
A process for holding operation information including at least element values of the physical system;
A process of storing the operation information in the operation information storage unit while collecting the operation information of the physical system;
A process of calculating the future element value of the physical system as a predicted element value based on the element value;
A prediction program for executing a process of predicting a future operating state of the physical system based on the prediction element value. In addition to the computer,
The prediction program according to claim 9, wherein processing for calculating the prediction element value is executed according to an autoregressive model.

Images