JP2005275558A - Energy load prediction device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device predicting a time-based delivery quantity at each delivery point in the day of supply for energy supply from a plurality of energy production lines to a plurality of delivery points through a pipeline network. <P>SOLUTION: A time-based delivery quantity correction means 21 corrects result data of time-base delivery quantity at each delivery point in one similar day selected based on characteristic information of the day of the supply so that result data of time-based production quantity of each energy production line are matched to a time-based total value. A second time-base delivery quantity calculation means 24 calculates a time-based delivery quantity in the day of supply by conduit network steady analysis by use of time-based production quantities obtained by decomposing a total demand prediction value for one day based on the time-base production quantity result of the similar day by a time-based total demand formation means 22. A first time-based delivery quantity calculation means 23 calculates a time-based delivery quantity in the similar day by conduit network steady analysis. A third time-based delivery quantity calculation means 25 adds the difference of time-based delivery quantity between the means 24 and means 23 to a time-base delivery quantity given by the correction means 21, and takes the result as a prediction value of time-based delivery quantity at each delivery point in the day of supply. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to the time at each of the payout bases on the supply day used to create an energy production supply plan for each production line and hour by which a plurality of energy production lines supply energy to a plurality of payout bases via a conduit network. The present invention relates to an energy load prediction apparatus that predicts a separate payout amount by computer calculation processing.

  As the most common form in which energy produced in a plurality of energy production lines is supplied to a plurality of dispensing bases via a conduit network, city gas (referred to as gas as appropriate) is assumed. Here, the production amount of the gas produced in the plurality of production lines is determined by the consumption amount of the consumer, and when the consumption amount of the consumer fluctuates, the production amount of each production line also fluctuates accordingly. The reason for this variation is that so-called pressure control (PC) operation is performed in which the pressure of the gas supply line from each production line to the consumer is adjusted to be constant. Therefore, when the consumer's consumption increases and the pressure decreases, the supply amount from each production line is immediately increased to adjust the pressure to be constant.

  The pressure of the gas supply line is divided between each production line and the customer (supply point) via high pressure, medium pressure, and low pressure conduit networks and high pressure and medium pressure regulators (governors). The pressure is reduced. Here, the high-pressure conduit network has a pressure of 1 MPa (in order to transport gas produced in a production line at a remote location from the customer for a long distance and in case the gas supply from each production line is stopped. It is kept at an extremely high pressure (megapascal) or higher. The specific operation pressure of the high-pressure conduit is set within an appropriate range according to the demand, owned equipment, and the like.

FIG. 2 is a block diagram schematically showing the gas supply system 100 from each production line to the consumer. In FIG. 2, first to fourth production lines 101 to 104 produce gas and deliver the gas to the high-pressure conduit network 110 with delivery amounts S1 to S4 (Nm ³ / h), respectively. The delivered gas is reduced in pressure by a plurality of high-pressure governors 105 to 108 on the conduit network 110, discharged to the intermediate-pressure conduit network 120, and further reduced in pressure from the intermediate-pressure conduit network 120 to be finally in demand. To a consumer (not shown) and consumed at a demand amount Z (Nm ³ / h). Therefore, the total amount of payouts to be paid out to the medium pressure conduit network 120 by the high pressure governors 105 to 108 is the demand amount Z.

In addition, a holder 130 is installed at one or more points in the vicinity of the installation locations of the high pressure governors 105 to 108 on the high pressure conduit network 110, and gas can be accumulated and released by opening and closing the valves. it can. With this function, when the consumer's consumption increases, the accumulated gas is released, and when the consumer's consumption decreases, the surplus gas is accumulated to mitigate the effects of sudden fluctuations in the consumer's consumption. To do. The holder 130 releases the gas with the discharge amount H (Nm ³ / h), and accumulates the gas when the value is negative. A similar gas consumption fluctuation adjustment function also includes the high-pressure conduit network 110 itself. This is because there is a line pack defined by the amount of gas that is stored under a given pressure in the entire geometric volume of the conduit network 110. By using the line pack, gas is supplied to the medium pressure conduit network 120 at a supply amount L (Nm ³ / h), and when the value is negative, the gas can be accumulated. Accumulation or supply of line packs means that the pressure in the high pressure conduit network 110 will increase or decrease, where the pressure in the conduit network 110 is not a constant value but a constant operating pressure range. Operation to keep in. Therefore, if the demand amount fluctuates, the pressure of the conduit network 110 fluctuates, and the line pack is supplied or accumulated, so that the fluctuation amount of the gas consumption can be absorbed within the operating pressure range. it can.

Here, the delivery amounts S1 to S4 (Nm ³ / h) from each production line, the discharge amount H (Nm ³ / h) of the holder 130, the line pack supply amount L (Nm ³ / h), and the medium pressure conduit network The relationship of the payout amount Z (Nm ³ / h) to 120 is expressed by the following formula 1.

(Equation 1)
S1 + S2 + S3 + S4 + H + L = Z

  However, although it is theoretically possible to control the pressure of the conduit network 110 to be constant by the holder 130 and the line pack, the number of the holders 130 is limited, and the actual number of the holder 130 and the line pack is limited. Since the total geometric volume is also relatively small, it is not practical to perform control to keep the pressure constant. Therefore, in order to adjust the pressure of the conduit network 110 to be constant or within the operating pressure range, the supply amount from each production line (corresponding to the fluctuation of the demand amount (dispensing amount Z to the conduit network 120 of medium pressure)) ( S1 + S2 + S3 + S4) needs to be adjusted.

  However, if the supply amount or the production amount of each production line varies, the production cost corresponding to the production amount also varies. Here, in order to explain the production cost in each production line, the gas production process will be outlined.

  First, the liquefied LNG taken out from the storage LNG tank by the LNG pump is vaporized and supplied by a vaporizer (for example, an open rack type vaporizer) using seawater pumped up by the seawater pump. The LNG gas vaporized in the LNG tank is supplied after being compressed by a BOG compressor. At this time, it may be further compressed by a BOG booster and supplied. The LNG gas supplied in this way is mixed with LPG separately taken out from the LPG tank by the LPG pump to adjust the amount of heat, and sent as city gas to the high-pressure conduit network. Here, it is normal that a plurality of production lines for sending out gas are provided for each production site, and there are cases where the presence or absence of heat amount adjustment and the delivery calorie value are different for each production line. .

  In the gas production process, power costs such as various pumps and compressors, heat adjustment costs, steam costs, water costs, and the like are intricately intertwined to determine gas production costs corresponding to the production amounts. For example, since a plurality of LNG pumps are usually provided and are sequentially activated as necessary, the operating cost varies depending on the number of activated units, that is, the delivery amount. As described above, the gas production cost varies depending on the amount of production.

  In the above, the actual supply amount (S1 + S2 + S3 + S4) from each production line does not immediately respond to fluctuations in demand in real time, but a 24-hour hourly production supply plan for the production supply day is created in advance. Gas supply from each production line is performed based on the production supply plan. For example, as disclosed in Patent Document 1 below, the total daily demand is predicted, the daily total demand predicted value is satisfied, and the above-described gas production cost is optimized. In addition, a production supply plan for each production line and each hour is created.

  Hereinafter, a conventional procedure for creating a production supply plan for each production line and each time will be described with reference to the flowchart of FIG.

  Based on the weather forecast (temperature, water temperature, weather, etc.) on the day of manufacturing supply and the characteristics information such as the day of the week and the actual demand data of the total demand on a similar day in the past, The daily total demand forecast value Z is derived (step # 100). Here, for example, one day is defined as 24 hours from 7 am on the current day to 7 am on the next day.

  Next, specify the most recent similar date whose characteristics information is similar to the date of manufacture supply date, and the actual data of the hourly payout amount from each high pressure governor on the similar date is obtained from each high pressure governor on the date of manufacture supply. Temporarily determined as hourly payout amount Z0 (i, j) (step # 101). Here, the argument i (i = 1 to 24) represents each time zone, for example, i = 1 is 7am to 8am on the day, and i = 24 is 6am to 7am on the next day. Each hour corresponds. The argument j represents the high pressure governor distinction.

  If it is difficult to determine the hourly payout amount Z0 (i, j) on one similar day, the production supply date is set, for example, from 7:00 am to 5:00 pm on that day. Divide into 3 hours from 5:00 pm to 10:00 pm, from 10:00 pm on the day to 7:00 am on the next day, select similar days for each time section, and from each high pressure governor on each similar day Actual data of hourly payout amount is provisionally determined as hourly payout amount Z0 (i, j) from each high-pressure governor on the day of manufacture and supply (step # 101 ').

  Next, the hourly payout amount so that the actual data of the hourly production amount on the similar day and the hourly total amount Z0 (i, j) of each temporarily determined high pressure governor coincide. Z0 (i, j) is corrected, and further, the corrected daily total amount is corrected so as to match the daily total demand amount predicted value Z to obtain the hourly payout amount Z1 (i, j). (Step # 102). Here, in the first correction, the difference between the hourly manufacturing amount actual data of the similar day and the hourly total amount of the hourly payout amount Z0 (i, j) is used as the hourly payout amount Z0 (i, j). Proportional distribution by high-pressure governor and adding to the hourly payout amount Z0 (i, j), with subsequent correction, dividing the daily total demand forecast value Z by the total production volume on the similar day Is multiplied by the total daily demand forecast value Z.

  Next, the daily total demand forecast value Z is decomposed for each production line and every hour based on the actual production data of the hourly production quantity for each production line on the same day, and the hourly production for each production line. A quantity prediction value S (i, k) is provisionally determined (step # 103). Here, the argument k represents the distinction between production lines. As described above, since the gas production cost fluctuates from production line to production line and time zone, the gas production cost is reduced by time so that the total gas production cost corresponding to the production amount of each production line is minimized at each hour. A temporary production amount predicted value S (i, k) for each production line is provisionally determined by a trial and error method such as distributing production amounts in order from a simple production line.

  Next, the manufacturing of each main node of the high-pressure conduit network is performed using the hourly payout amount Z1 (i, j) corrected in step # 102 and the temporarily determined hourly manufacturing amount predicted value S (i, k) as input conditions. A pressure transition for 24 hours on the supply day is derived by simulation using a conduit network transient response analysis tool (step # 104). Here, the pipe network transient response analysis tool is a tool for deriving the transient response of the compressible fluid in the high-pressure gas pipeline network by unsteady analysis. It is assumed that the high-pressure conduit network to be simulated is modeled in advance including the holder.

  Next, it is confirmed whether the pressure transition obtained in step # 104 is within a predetermined operating pressure range (step # 105). Here, when the pressure transition is outside the operating pressure range, the temporarily determined production amount predicted value S (i, k) for each production line is changed (step # 106). Here, in the change of the hourly production amount prediction value S (i, k), adjustment is performed so that the total gas production cost corresponding to the production amount of each production line is minimized at each time. Then, the derivation of the pressure transition in step # 104 and the confirmation in step # 105 are repeated until the pressure transition falls within the predetermined operating pressure range.

  In Step # 105, when it is confirmed that the pressure transition is within the predetermined operating pressure range, the production supply is performed based on the hourly production amount predicted value S (i, k) for each production line that has been provisionally determined or changed. An hourly production supply plan for each production line for 24 hours on the day is created (step # 107).

  On the date of manufacture and supply, on the basis of the hourly manufacture and supply plan for each production line, the production amount change command and execution monitoring for each production line are performed for each manufacturing site (step # 108).

  After manufacturing supply based on the hourly production supply plan for each production line created on the day of production supply date, the hourly total production amount forecast value S (i) and the actual data of hourly total demand ( If the difference from the actual data of the hourly payout amount from each high pressure governor exceeds a predetermined value (step # 109), the manufacturing is started from step # 101 in order to review the manufacturing supply plan. Run supply planning again.

Further, it is determined whether or not the pressure measurement value at a predetermined point on the high-pressure conduit network exceeds a predetermined operating pressure range (step # 110). A production amount change command is issued so that the pressure at the point falls within the operating pressure range (step # 111), and production and supply plan creation is executed again from step # 101.
JP 2002-334138 A

  The inventor of the present invention can correct the amount of discharge from the high pressure governor so as to balance the amount of production supplied to the high pressure conduit network, the amount of discharge / accumulation in the line pack, and the amount of discharge from the high pressure governor. If the estimated daily demand, which is a precondition, matches, we have empirically confirmed that the results of the unsteady analysis using the pipeline network transient response analysis tool and the actual results are almost the same. However, in the above-described conventional production line-by-hour and hour-by-hour production / supply plan creation method, the correction process itself is simple in the method for deriving the hourly payout amount at each payout base on the supply day (steps # 101 to # 102). Although there is a possibility that the operation state of the line pack of the high-pressure conduit network for the hourly payout amount Z0 (i, j) temporarily determined in step # 101 may be changed by the correction process in step # 102. The hourly payout amount at each payout base after correction may not reflect the exact load distribution on the conduit network.

  In addition, when changing the operating state of the high-pressure governor on the supply day from the operating state of the high-pressure governor on the similar date, or when changing the operating state of the conduit network including high-pressure, The amount to be paid out was corrected based on experience with the amount determined from experience.

  Therefore, based on the separately estimated daily demand, simplification of similar days based on weather conditions such as temperature and water temperature is simplified, and each high pressure is calculated by correcting the hourly discharge amount (high pressure governor load) from each high pressure governor. A production line that satisfies the operating pressure range of the high-pressure conduit network for the preceding 24 hours by automating the determination of hourly payout amount of the governor, changing the operation of the factory, and inputting priorities for various efficiency improvement conditions. There has been a demand for a system that can quickly and automatically calculate the amount of production per hour and create a production supply plan.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to solve the above-described problems and allow a plurality of energy production lines to supply energy to a plurality of payout bases via a conduit network. It is an object of the present invention to provide an energy load prediction device capable of automatically creating hourly payout amounts at each payout base on the day of supply used for creating energy production supply plans for each production line and hour by computer calculation processing.

  In order to achieve this object, the first characteristic configuration of the energy load predicting apparatus according to the present invention is that a plurality of energy production lines supply energy to a plurality of payout bases via a conduit network according to production line, time. An energy load prediction device for predicting hourly payout amounts at the respective payout bases on the supply day used for creating another energy production supply plan by a computer calculation process, the past hourly production amounts of the respective energy production lines, An input database for receiving an input of a past result hourly payout amount and an hourly secondary pressure actual data at each payout base, and a predicted daily total demand amount predicted on the supply day separately And selecting one similar date based on the characteristic information of the supply day, and before each payout base on the selected similar date Hourly payout amount correction that corrects the hourly payout amount actual data so that the hourly total value matches the hourly total value of the hourly production amount actual data of each energy production line on the similar day. And a time for decomposing the daily total demand forecast value into hourly total demand forecast values on the supply day based on actual data of hourly production quantities of the energy production lines on the selected similar days. Separate total demand creation means, actual data of hourly production amount of each energy production line on the similar date, and hourly operation result data of secondary pressure at each payout base as at least input conditions, A first hourly payout amount calculating means for calculating an hourly payout amount at each payout base on a similar date, an hourly total demand forecast value of each energy production line on the supply day, and A second hourly payout amount calculating means for calculating an hourly payout amount at each payout site on the supply day, with at least an hourly operation plan of secondary pressure at each payout base as an input condition; and the hourly payout From the hourly payout amount at each payout base on the supply day calculated by the second hourly payout amount calculation means to the actual data of the hourly payout amount at each payout base after correction by the amount correcting means A value obtained by subtracting the hourly payout amount at each payout base on the similar date calculated by the first hourly payout amount calculating means for each payout base is added for each payout base, and each payout base on the supply day And a third hourly payout amount calculation means which is a predicted value of the hourly payout amount.

  In the second feature configuration, in addition to the first feature configuration, the first hourly payout amount calculation means and the second hourly payout amount calculation means calculate the hourly payout amount at each payout base. And a conduit network steady state analysis tool for analyzing a conduit network for steady state analysis including the conduit network and a second conduit network having a lower pressure than the conduit network connected to the conduit network via the payout bases. There is in point to use.

  In the third feature configuration, in addition to the first or second feature configuration, the hourly payout amount correcting unit is configured to change the hourly payout record data of the hourly payout amount at each payout base on the similar date. The difference between the total value and the hourly total value of the hourly production amount actual data of each energy production line on the similar day is configured as the hourly payout amount actual data of the respective payout bases on the similar day. It is prorated by the ratio and is added to the actual data of the hourly payout amount of each payout base on the similar day.

  According to the first to third characteristic configurations of the energy load predicting apparatus, first, similar days having similar characteristics information (meteorological conditions and day of the week) on the supply day are selected by the hourly payout amount correcting means, and further manufactured. The actual data of the hourly payout amount of each payout base on a similar day is corrected so that the mass balance between the amount and the payout amount can be achieved. Further, in order to correct the actual data of the hourly payout amount of each payout base on the similar day corrected by the hourly payout amount correcting means, the first data is corrected to be consistent with the daily total demand amount forecast value on the supply day. Use the difference between the hourly payout amount at each payout base on a similar day calculated by the hourly payout amount calculation means and the hourly payout amount at each payout base on the supply day calculated by the second hourly payout amount calculation means. Therefore, it is possible to make corrections in line with the actual operational state of the line pack on the conduit network, and finally the predicted value of hourly payout amount at each payout base on the supply day calculated by the third hourly payout amount calculating means. As a result, a high-accuracy product can be obtained as compared with the prior art. As a result, energy production and supply plans for each production line and hour can be created with high accuracy.

  In particular, according to the second characteristic configuration, a more accurate value can be obtained as the predicted value of the hourly payout amount at each payout base on the supply day. In addition, if there is a change in the connection state of the conduit network due to the construction of the conduit on the day of supply as a steady analysis conduit network, a steady state analysis conduit network reflecting the change in the connection state should be created and used. Thus, it is possible to obtain a predicted value of the hourly payout amount at each payout base more in line with the actual operation. Further, according to the third feature configuration, the correction by the hourly payout amount correcting means can be executed easily and reliably, and the mass balance between the production amount and the payout amount on the similar date can be taken.

  In the present invention, the term “by hour” is not particularly limited to one hour unit, and may be, for example, every 30 minutes or every two hours.

  The fourth feature configuration is that, in addition to any one of the feature configurations described above, the input means accepts an input of an hourly operation plan of secondary pressure at each payout base or a change thereof.

  According to the fourth characteristic configuration of the energy load prediction apparatus, an hourly operation plan of the secondary pressure at each payout base is adopted, and an hourly payout amount at each payout base on the supply day is predicted. A value can be obtained. This is because if a similar date is selected in consideration of the similarity of hourly operation plans at the payout base, an appropriate similar date may not necessarily be selected. Can be avoided.

  In the fifth feature configuration, in addition to any one of the feature configurations described above, the performance database further stores performance data of hourly discharge amounts of one or more holders on the conduit network, The hourly payout amount calculation means further uses the actual data of the hourly discharge amount of the holder as the input condition, and the second hourly payout amount calculation means further determines the hourly operation plan of the holder as the input condition. It is in the point to.

  According to the fifth characteristic configuration of the energy load prediction device, it is possible to obtain a more accurate predicted value of hourly payout amount at each payout base on the supply day in consideration of a case where a holder is installed on the conduit network. it can.

  The sixth feature configuration is that, in addition to the fifth feature configuration, the input means accepts an input of the hourly operation plan of the holder or its change.

  According to the 4th characteristic structure of the said energy load prediction apparatus, the arbitrary value is employ | adopted as an hourly operation plan of a holder, and the predicted value of the hourly payout amount in each payout base on the supply day can be obtained. This is because if a similar date is selected in consideration of the similarity of the holder's hourly operation plan, an appropriate similar date may not necessarily be selected. However, according to this feature configuration, such inconvenience can be avoided in advance. it can.

  An embodiment of an energy load prediction apparatus according to the present invention (hereinafter referred to as “the present invention apparatus” as appropriate) will be described with reference to the drawings. In the present embodiment, the apparatus of the present invention has a high pressure in a plurality of production lines 101 to 104 (corresponding to energy production lines) in the gas supply system 100 illustrated in FIG. 2 assuming city gas as the energy type. From a plurality of high-pressure governors 105-108 (corresponding to the payout base) via the conduit network 110 to the medium-pressure conduit network 120 connected to the high-pressure governors 105-108, a city with a production amount that matches the total demand on the day of supply This is a support system for supplying gas, and generates hourly payout amounts in each high-pressure governor on the supply day used for preparation of energy production supply plans for each production line and hour.

  As shown in FIG. 1, the device 1 of the present invention includes an hourly load prediction unit 20 that creates an hourly discharge amount (hourly load prediction value) in each high pressure governor on the supply day, and one day of the separately predicted supply day. Input means 2 that accepts input of total demand forecast value, etc., actual data of hourly production amount of each production line 101-104, hourly operation result in each high pressure governor 105-108 (payout amount, primary side pressure Secondary pressure, etc.), a performance database 4 for storing hourly operation results (discharge amount, accumulated amount, etc.) of the holder 130, and the like. In addition, the supply day does not necessarily mean one day according to the calendar as described above, and can be changed as appropriate.

  Furthermore, the device 1 of the present invention is a user terminal 3 used by the user (operator) of the device 1 of the present invention for input / output operations, and optimizes manufacturing costs for 24 hours on the supply day (for example, from 7:00 am Manufacturing supply plan creation unit 10 that creates hourly production quantities for each production line at 7:00 am on the following day, weather database 5 that stores weather data (temperature, water temperature, weather, etc.) on the past production date, high pressure and medium pressure Via a conduit network database 6, a conduit network transient response analysis tool 7, a conduit network steady state analysis tool 8, and a computer network 9 such as a LAN. Connect. A part of the data input to the input means 2 is used in common by the device 1 of the present invention and the production / supply plan creation unit 10.

  The hourly load prediction unit 20 of the device 1 of the present invention further includes an hourly payout amount correcting means 21, an hourly total demand amount creating means 22, a first hourly payout amount calculating means 23, and a second hourly payout amount calculating means. 24 and third hourly payout amount calculation means 25. The production supply plan creation unit 10 includes a supply pattern generation unit 11, a pressure transition determination unit 12, a production cost calculation unit 13, an optimum supply pattern derivation unit 14, an error output unit 15, and a pressure abnormality prediction unit 16. Composed. Here, in the present embodiment, the input unit 2, the manufacturing and supply plan creation unit 10, and the hourly load prediction unit 20 are realized by executing the processing of each unit in software on computer hardware.

  Next, prediction processing of hourly payout amount Z1 (i, j) in each high-pressure governor on the supply day using the apparatus 1 of the present invention, production supply plan creation procedure by the production supply plan creation unit 10, and production supply plan creation The functions and operations of the units of the unit 10 and the hourly load prediction unit 20 will be described with reference to the flowchart shown in FIG.

  First, the user accesses the weather forecast (temperature, water temperature, weather, etc.) on the day of supply at a user terminal by accessing a weather data server (not shown) that provides external weather forecast data and obtains the weather forecast. Based on the weather forecast data and the day of the supply day (both are collectively referred to as “characteristic information”), the daily total demand amount on the supply day is predicted, and the daily total demand amount predicted value Z is It inputs into the input means 2 of this invention apparatus 1 with information (step # 1). Here, the daily total demand forecast value Z is a correspondence relationship between the past daily total demand stored in the performance database 4 in advance and the weather conditions (temperature, water temperature, weather, etc.) on each supply date. , For example, by deriving a primary regression equation etc. with the supply day of the week and each weather condition as explanatory variables and the daily total demand forecast value Z as an objective variable, based on the regression equation The daily total demand forecast value Z is obtained.

  When the daily total demand forecast value Z is input to the input means 2, the hourly load prediction unit 20 is activated, and the hourly payout amount Z1 (i, j) in each high-pressure governor on the supply day in the processing procedure described later. (Corresponding to the hourly load prediction value) is predicted (step # 10). Here, the argument i (i = 1 to 24) represents each time zone, for example, i = 1 is 7am to 8am on the day, and i = 24 is 6am to 7am on the next day. Each hour corresponds. The argument j represents the high pressure governor distinction.

  Next, the production and supply plan creation unit 10 optimizes the production cost on the day of supply based on the daily total demand forecast value Z and the hourly payout amount Z1 (i, j). A quantity S (i, k) is obtained (step # 20). Here, the argument k represents the distinction between production lines. Hereinafter, the prediction process of the hourly payout amount Z1 (i, j) in each high-pressure governor on the supply day of step # 10 will be described in detail.

  First, the hourly payout amount correction means 21 accesses the weather database 5 and is most similar to the supply day based on the supply day characteristic information (day of week, temperature, water temperature, weather, etc.) received by the input means 2. One similar date is selected, and the actual hourly payout data Z0 (i, j) of each high-pressure governor selected for the selected similar date, the hourly total value Z0 (i) is the production line for each similar date. Is corrected so as to coincide with the hourly total value S0 (i) of the actual production data S0 (i, k) of the hourly production amount (step # 11). Specifically, as shown in Equation 2 below, the hourly payout amount correction unit 21 calculates the hourly total amount Z0 (i) of the hourly payout amount of each high pressure governor on the selected similar day, and The difference (S0 (i) -Z0 (i)) with the hourly total value S0 (i) of the actual hourly production amount of each production line on the similar day is used as the hourly payout amount of each high-pressure governor on the similar day. The actual data Z0 (i, j) is proportionally distributed and added to the actual data Z0 (i, j) of the hourly payout amount of each high-pressure governor. In Equation 2, Z0 '(i, j) is the hourly payout amount of each high-pressure governor after the correction processing in Step # 11.

(Equation 2)
Z0 '(i, j) = Z0 (i, j)
+ (S0 (i) −Z0 (i)) × Z0 (i, j) / Z0 (i)

  Next, the hourly total demand amount creating means 22 uses the daily total demand amount predicted value Z received by the input means 2 as the actual data S0 of hourly production quantity of each production line selected in step # 11. Based on (i, k), the hourly total demand forecast value Z (i) on the supply day is decomposed as shown in the following Equation 3 to generate the hourly total demand forecast value Z (i). (Step # 12). In Equation 3, S0 is a total sum value for one day of the hourly total value S0 (i) of the record data S0 (i, k).

(Equation 3)
Z (i) = Z × S0 (i) / S0

  Next, the first hourly payout amount calculating means 23 records actual data S0 (i, k) of hourly production amount of each production line selected on the similar date selected in Step # 11, 2 in each high pressure governor on the similar date. The hourly operation result data PS0 (i, j) of the secondary pressure and the result data H0 (i, m) of the hourly discharge amount of each holder 130 on the similar day (m indicates the distinction between the holders). As an input condition, the hourly payout amount Z0s (i, j) in each high-pressure governor on the similar date is calculated (step # 13). Specifically, the first hourly payout amount calculation means 23 includes a high-pressure conduit network 110 and a medium-pressure conduit network 120 connected to the high-pressure conduit network 110 via each high-pressure governor (corresponding to a second conduit network). ) Is extracted from the pipeline network database 6 and each of the network on the steady analysis conduit network on the similar date is extracted from the pipeline network database 6 using the pipeline network steady analysis tool 8. The hourly discharge amount Z0s (i, j) from the high-pressure conduit network 110 to the medium-pressure conduit network 120 in the high-pressure governor is obtained by simulation. In the steady analysis conduit network, various components (high pressure governors, holders, conduits, etc.) in the high pressure and medium pressure conduit networks 110 and 120 are preliminarily modeled and stored in the conduit network database 6. It shall be.

  Next, the second hourly payout amount calculation means 24 performs the hourly total demand forecast value Z (i) for each production line on the supply day generated in step # 12, the secondary side in each high pressure governor on the supply day. With the hourly operation plan PS1 (i, j) of the pressure and the hourly operation plan H1 (i, m) of the holder on the supply day as input conditions, the hourly discharge amount Z1s (i, j) is calculated (step # 14). Specifically, the second hourly payout amount calculating means 24 extracts the same steady analysis conduit network used in step # 13 from the conduit network database 6, and the conduit network steady state is extracted from the steady analysis conduit network. The analysis tool 8 is used to simulate the hourly discharge amount Z1s (i, j) from the high pressure conduit network 110 to the intermediate pressure conduit network 120 in each high pressure governor on the steady analysis conduit network on the supply day. Ask.

  Next, the third hourly payout amount calculating means 25 sets the second hourly payout amount calculating means to the hourly payout amount Z0 ′ (i, j) of each high-pressure governor corrected by the hourly payout amount correcting means 21. 24. Hourly payout amount Z0s (for each high-pressure governor on the similar day calculated by the first hourly payout amount calculation means 23 from the hourly payout amount Z1s (i, j) for each high-pressure governor on the supply day calculated by 24. The value obtained by subtracting i, j) for each high-pressure governor and each hour is added to each high-pressure governor and each hour to obtain a predicted value Z1 (i, j) of hourly payout amount in each high-pressure governor on the supply day ( Step # 15). In summary, Z1 (i, j) is calculated by the following equation (4).

(Equation 4)
Z1 (i, j) = Z0 ′ (i, j) + Z1s (i, j) −Z0s (i, j)

  Next, based on the daily total demand forecast value Z and the hourly payout amount Z1 (i, j) obtained in Step # 10 (Steps # 11 to # 15) by the manufacturing and supply plan creation unit 10, The hourly production quantity S (i, k) for each production line that optimizes the production cost on the supply day is obtained. Hereinafter, the processing procedure will be described.

  The supply pattern generation unit 11 uses a genetic algorithm in steps # 21, # 22, and # 26, which will be described later, and includes a supply pattern composed of hourly production quantities S (i, k, l) for each production line on the supply day. Generate multiple patterns. Here, the argument l represents the generated supply pattern.

  First, the supply pattern generation means 11 accesses the weather database 5 and selects one or a plurality of similar days (for example, for 1 to 30 days) based on the characteristic information (day of week, temperature, water temperature, weather, etc.) on the supply day. Hourly production for each production line on the day of supply based on the actual data of hourly production volume for each production line selected and each day selected and the predicted daily total demand amount Z received by the input means 2 A plurality of supply patterns (for example, 100 to 10000 patterns) including the amount S (i, k, l) are generated and set as the first generation pattern (step # 21). For example, in creating the hourly production quantity S (i, k, l) in step # 21, first, the daily total demand forecast value Z is calculated from the hourly total production quantity on each selected similar day (each production line). Based on the actual data of the hourly production amount), and is distributed to the hourly total demand forecast value Z (i, l) based on the hourly composition ratio, and the hourly production amount S of each production line The constraint is that the sum of (i, k, l) is the hourly total demand forecast value Z (i, l). Moreover, in each supply pattern of the 1st generation, when distributing the hourly total demand predicted value Z (i, l) to each production line, the performance data of each similar day is referred to. Therefore, each supply pattern of the 1st generation is not generated at random, but is generated according to the performance data of similar days.

  Next, for each newly generated supply pattern of a certain generation (for example, the first generation), the pressure transition determination means 12 performs an hourly manufacturing amount S (i, k, l) and hourly discharge amount Z1 (i, j) in each high pressure governor, as input conditions, the pressure transition (temporal change in pressure) of one or a plurality of predetermined nodes of the high pressure conduit network 110 for each supply pattern, It is derived using the conduit network transient response analysis tool 7 (corresponding to a conduit network unsteady simulator), and it is determined whether or not the derived pressure transition is within a predetermined operating pressure range (step # 22). Here, as the predetermined node, for example, an intermediate point on the high-pressure conduit network, a city gas delivery point of a plurality of production lines, or the like is used. And the said determination is performed using the operating pressure range defined for every predetermined node. Here, the conduit network transient response analysis tool 7 is a tool for deriving the transient response of the compressible fluid in the high-pressure gas pipeline network by unsteady analysis, and the high-pressure conduit network to be simulated is stored in the conduit network database. 6 is extracted and used. It is assumed that the simulation conduit network including the holder is modeled in advance and stored in the conduit network database 6.

  Further, in parallel with step # 22, the manufacturing cost calculation means 13 applies the hourly manufacturing amount S (for each manufacturing line) for each newly generated supply pattern of a certain generation (for example, the first generation). Based on i, k, l), the manufacturing cost for each manufacturing line and the total manufacturing cost are calculated (step # 23). The production cost for each production line is the production cost due to the difference in production cost due to the difference in power unit price by time for each production line, the presence or absence of power generation facilities, the number of pumps used, the combination of use, etc. Taking into account the fluctuation factors of the manufacturing cost, such as changing depending on the gas delivery amount, the calculation cost is calculated for each hour using a predetermined calculation formula, and the manufacturing cost for each manufacturing line for 24 hours is calculated. The calculated manufacturing cost for each manufacturing line is totaled to obtain the daily manufacturing cost.

  Next, the supply pattern generation unit 11 is a supply pattern in which the pressure transition of the predetermined node is determined to be within the predetermined operating pressure range by the pressure transition determination unit 12 from the generated generation patterns for one generation. In addition, a predetermined number (for example, 50 to 5000 patterns) is selected with the supply patterns having a low daily manufacturing cost calculated by the manufacturing cost calculating means 13 as excellent supply patterns (step # 24).

  Next, it is determined whether or not a predetermined number of supply patterns have been generated, or whether a predetermined convergence condition is satisfied (step # 25). If the end condition is not satisfied, Furthermore, a plurality of next-generation supply patterns are generated.

  The supply pattern generation means 11 performs a crossover process and a mutation process using a genetic algorithm on the selected excellent supply pattern, and supplies the next generation pattern (corresponding to the second generation if it is the first time). Is generated (step # 26). When a new generation supply pattern is generated, steps # 22 to # 25 are repeated until the end condition of step # 25 is satisfied.

  When the end condition is satisfied in the determination process of step # 25, the error output means 15 then sets the predetermined node by the pressure transition determination means 12 among the supply patterns of the respective generations generated in steps # 21 to # 26. It is determined whether the absolute number or the ratio of the supply patterns determined that the pressure transition is within the predetermined operating pressure range exceeds the predetermined number (step # 27).

  If the number of supply patterns (absolute number or ratio) determined that the pressure transition of the predetermined node is within the predetermined operating pressure range in step # 27 exceeds the predetermined number, the predetermined operating pressure range Is determined to be appropriate, and a supply pattern having the minimum daily production cost is extracted from all of the generated excellent supply patterns of all generations (step # 28).

  Further, when the error output means 15 determines in step # 27 that the number of supply patterns (absolute number or ratio) determined that the pressure transition of the predetermined node is within the predetermined operating pressure range is equal to or less than the predetermined number. Extracts a node whose pressure transition is outside the predetermined operating pressure range for the supply pattern that does not satisfy the above-described determination condition among the supply patterns of each generation generated in steps # 21 to # 26, Information about the node is output as a message to the user terminal 3 (step # 29).

  As described above, the production and supply plan creation unit 10 optimizes the production cost on the supply day based on the daily total demand amount predicted value Z and the hourly payout amount Z1 (i, j) by the processes in steps # 21 to # 28. When the hourly production quantity S (i, k) for each production line to be converted is obtained, a production supply plan on the supply day is created based on the hourly production quantity S (i, k) for each production line, Based on the plan, a manufacturing instruction is given to each manufacturing line (step # 30).

  When production of each production line is executed based on the created production supply plan, the production volume of each production line, the pressure of a predetermined node on the conduit network, the operating state of the high-pressure governor, etc. are sequentially measured, and the results database 4 is registered as new data. At the same time, it is monitored whether the actual data at the time of execution is inconsistent with the planned value or does not deviate from the predetermined operation range (step # 31). Thus, before the production on the supply day ends, the production supply plan for the next supply day is created by returning to step # 1 and repeating.

  In the execution monitoring of step # 31, the pressure abnormality predicting means 16 sequentially accesses the result database 4, acquires the result data of the hourly pressure transition of each of the predetermined nodes on the supply day, and from the result data of the hourly pressure change. Estimate the pressure transition of the preceding predetermined time. For example, as schematically shown in FIG. 4, when the historical data of the hourly pressure transition of each predetermined node has a downward trend, the downward curve DL (or straight line) of the hourly pressure transition is a straight line or a quadratic curve. Therefore, the descending curve DL is calculated for each predetermined node based on the acquired result data. Next, the pressure of the calculated descending curve DL deviates from the predetermined operating pressure range and the time (time) when the deviation from the calculated value of the pressure of each predetermined node derived by the pressure transition determination means 12 exceeds the predetermined value. Each time (time) is predicted. If the predicted time is within a predetermined time (for example, 3 hours), the error output means 15 issues an alarm and sets the predicted time or predicted time to the content (node distinction, deviation from the calculated value). A message is output to the user terminal 3 along with the deviation of the operating pressure range. Upon receiving the message, the user re-creates the production supply plan and instructs the production line to change the necessary production amount.

  Hereinafter, another embodiment will be described.

  <1> In the above embodiment, the input unit 2 further includes the conduit network as the predetermined operating pressure range used by the pressure transition determination unit 12 to determine the pressure transition of one or more predetermined nodes of the high-pressure conduit network 110. It is also preferable that at least one of the minimum operating pressure of a predetermined node on 110 and the operating pressure range of the supply pressure for each production line can be received.

  In this case, the pressure transition determination means 12 preferentially uses the operating pressure range input with respect to the preset operating pressure range, so that the user of the device 1 of the present invention can operate the operating pressure range as necessary. Can be changed arbitrarily.

  Further, when the predetermined operating pressure range used by the pressure transition determining unit 12 is not set in advance and needs to be input every time, the pressure transition determining unit 12 determines whether the predetermined node input to the input unit 2 The operating pressure range may be used.

  <2> In the above embodiment, the input unit 2 may be configured to accept at least one input of a plurality of cost calculation conditions used by the manufacturing cost calculation unit 13 to calculate the manufacturing cost in each manufacturing line. preferable. The manufacturing cost calculation means 13 preferentially uses the cost calculation conditions input with respect to the preset cost calculation conditions, so that the user of the device 1 of the present invention can use the calculation costs as necessary. A plurality of cost calculation conditions can be arbitrarily changed.

  Further, when the plurality of cost calculation conditions used by the manufacturing cost calculation unit 13 are not set in advance and need to be input each time, the manufacturing cost calculation unit 13 inputs the cost calculation condition input to the input unit 2. May be used.

  <3> In the above-described embodiment, the input means 2 further includes an operation ratio of a manufacturing facility including a manufacturing line or a plurality of parts of the manufacturing line, or a daily production amount or a fixed amount in a specific one or more manufacturing lines. It is also preferable to be able to accept the input of the hourly production amount. When the supply pattern generation means 11 generates the hourly production quantity S (i, k, l) for each production line on the supply day in steps # 21 and # 26, the above operation input to the input means 2 The ratio, daily production amount, or fixed hourly production amount is set as a constraint. Therefore, when creating the first generation supply pattern in step # 21, normally, the time of each production line is based on the actual production data of the hourly production amount in each production line on each selected similar day. The allocation of the hourly production quantity between production lines is determined on the condition that the total of the different production quantities S (i, k, l) becomes the hourly total demand forecast value Z (i, l). When the above operation ratio, daily production amount or fixed hourly production amount is input to the input means 2, the distribution of the hourly production amount between the production lines is determined further bound by the inputted conditions. .

  For example, if the operation ratio of the production line 101 in FIG. 2 is set to 25%, the remaining 75% is allocated to other production lines based on the actual data. Further, for example, when the production lines 101 and 102 belong to the same factory A, and the operation ratio of the factory A is set to 40%, the production amount of 40% per hour in the production lines 101 and 102 is the above record. Based on the data, the remaining 60% of the hourly production amount is distributed on another production line based on the actual data.

  Further, for example, when the daily production amount of the production line 101 in FIG. 2 is set, the daily production amount of the production line 101 from the hourly production amount S (i, k, l) obtained by the normal processing procedure. And the hourly production quantity S (i, k, l) of the production line 101 is corrected at the same ratio so that the calculated daily production quantity becomes the set daily production quantity. Differences according to time caused by the correction are distributed on other production lines.

  Further, for example, when the production amount by fixed time of the production line 101 in FIG. 2 is set, the production amount by hour S (i, k, l) of the production line 101 is fixed to the set production amount by fixed time. The sum of the hourly production quantities S (i, k, l) of the other production lines is a value obtained by subtracting the production quantity by fixed time from the hourly total demand forecast value Z (i, l). With this as a new constraint condition, hourly production quantities S (i, k, l) of other production lines are generated.

  <4> In the above-described embodiment, the second hourly payout amount calculation unit 24 uses the hourly operation plan PS1 (i, j) of the secondary pressure in each high pressure governor and the hourly operation plan H1 (i, m) of the holder. ) Describes the case where a normal operation plan (default value) set in advance is used, but the input means 2 further includes an hourly operation plan PS1 (i, j) or a holder of the secondary pressure in each high pressure governor. It is also preferable that the hourly operation plan H1 (i, m) or both of these inputs or their change inputs can be received. In this different embodiment, the second hourly payout amount calculation means 24 uses the input means 2 when calculating the hourly payout amount Z1s (i, j) in each high pressure governor by using the conduit network steady state analysis tool 8. The hourly operation plan PS1 (i, j) of the secondary pressure inputted from the above or the hourly operation plan H1 (i, m) of the holder is preferentially used as an input condition. The second hourly payout amount calculation means 24 corresponds to the hourly operation plan PS1 (i, j) of the secondary pressure in each high pressure governor set in advance or the hourly operation plan H1 (i, m) of the holder. By using the input operation plan preferentially, the user of the device 1 of the present invention can use the operation plan for simulating the hourly payout amount Z1s (i, j) in each high-pressure governor on the supply day as necessary. Can be changed arbitrarily.

  Further, when the operation plans used by the second hourly payout amount calculating means 24 are not set in advance and need to be input each time, the second hourly payout amount calculating means 24 inputs to the input means 2. The operation plans PS1 (i, j) and H1 (i, m) may be used.

  <5> In the above-described embodiment, the manufacturing cost calculation unit 13 performs, in step # 23, in parallel with step # 22, for each newly generated supply pattern of a certain generation (for example, the first generation). In the above description, the manufacturing cost for each manufacturing line and the total manufacturing cost are calculated. However, step # 23 may be executed after step # 22. That is, in step # 22, the manufacturing cost for each manufacturing line and the total manufacturing cost value may be calculated for the supply pattern in which the derived pressure transition is determined to be within the predetermined operating pressure range.

  <6> In the above-described embodiment, the manufacturing cost is optimized and the hourly production amount S (i, k) for each production line for 24 hours on the supply day (for example, 7 am to 7 am on the next day) A case has been described in which prediction is made based on the hourly payout amount Z1 (i, j) in each high-pressure governor for 24 hours on the same supply day. In this case, the hourly payout amount Z1 (i, j) is calculated for the 24-hour period. However, when various performance data, weather forecast data, etc. have a daily break of midnight, the hourly payout amount. Z1 (i, j) may be derived for 48 hours, and the necessary 24 hours may be extracted and used to create a manufacturing and supply plan.

  <7> In the above embodiment, city gas is assumed as the energy type, but the energy type is not limited to city gas as long as it is supplied using a conduit network. For example, the energy type may be water, petroleum, hot water, hydrogen, or the like.

The block diagram which shows one Embodiment of the energy load prediction apparatus which concerns on this invention, and its peripheral device The block diagram which represented typically the gas supply system used as the object of the energy load prediction apparatus which concerns on this invention The flowchart which shows an example of the preparation procedure of the hourly payout amount in each payout base on the supply day using the energy load prediction device according to the present invention, and the manufacturing supply plan The figure explaining the estimation process of the pressure transition of the preceding predetermined time by a pressure abnormality prediction means A flowchart showing a conventional procedure for creating a production supply plan for each production line and each hour

Explanation of symbols

1: Energy load prediction device according to the present invention 2: Input means 3: User terminal 4: Performance database 5: Meteorological database 6: Pipe network database 7: Pipe network transient response analysis tool 8: Pipe network steady state analysis tool 9: Computer Network 10: Production supply plan creation unit 11: Supply pattern generation means 12: Pressure transition determination means 13: Manufacturing cost calculation means 14: Optimal supply pattern derivation means 15: Error output means 16: Pressure abnormality prediction means 20: Hourly load prediction Unit 21: Hourly payout amount correcting means 22: Hourly total demand amount creating means 23: First hourly payout amount calculating means 24: Second hourly payout amount calculating means 25: Third hourly payout amount calculating means 100: Gas supply system 101-104: Production line (energy production line)
105-108: High-pressure governor (dispensing base)
110: High pressure conduit network 120: Medium pressure conduit network 130: Holder

Claims (6)

The amount of hourly payout at each of the payout bases on the supply day used to create an energy production supply plan for each production line and hourly for supplying energy to a plurality of payout bases via a conduit network. An energy load prediction device for predicting by computer processing,
A past record database storing past past hourly production quantities of the respective energy production lines, past past hourly payout quantities at the respective payout bases, and past hourly secondary pressure data;
An input means for receiving an input of a daily total demand forecast value predicted separately on the supply day;
One similar day is selected based on the characteristic information on the supply day, and the actual data of the hourly payout amount of each payout base on the selected similar day is the total value by hour, Hourly payout amount correction means for correcting the energy production line so as to match the hourly total value of the actual production amount of each energy production line,
The hourly total demand that decomposes the daily total demand forecast value into the hourly total demand forecast value of the supply day based on the actual data of hourly production quantity of each energy production line on the selected similar day A quantity creation means;
The actual data of hourly production amount of each energy production line on the similar date, and the hourly operation result data of secondary pressure at each payout site at least as input conditions, at each payout site on the similar date A first hourly payout amount calculating means for calculating an hourly payout amount;
Time at each payout site on the supply day, with at least the hourly total demand forecast value for each energy production line on the supply day and the hourly operation plan of the secondary pressure at each payout site as input conditions A second hourly payout amount calculating means for calculating a separate payout amount;
In the actual data of the hourly payout amount of each payout base after the correction by the hourly payout amount correcting means, the hourly payout base on the supply day calculated by the second hourly payout amount calculating means. A value obtained by subtracting the hourly payout amount at each payout base on the similar date calculated by the first hourly payout amount calculation means from the payout amount for each payout base is added for each payout base, and the supply day A third hourly payout amount calculation means as a predicted value of hourly payout amount at each payout base;
An energy load prediction apparatus comprising:   The first hourly payout amount calculating means and the second hourly payout amount calculating means calculate the hourly payout amount at each payout base by using the conduit network, the conduit network, and the payout bases. 2. The energy load prediction apparatus according to claim 1, wherein a conduit network steady analysis tool for analyzing a steady analysis conduit network including a second conduit network having a pressure lower than that of the conduit network connected thereto is used.   The hourly payout amount correcting means includes an hourly total value of the hourly payout amount actual data of the respective payout bases on the similar date and the hourly production amount actual data of the energy production lines on the similar day. Is divided by the composition ratio of the actual data of the hourly payout amount of each payout base on the similar date, and the difference of the hourly total value of The energy load prediction apparatus according to claim 1, wherein the energy load prediction apparatus adds the result data separately to each other.   The energy load prediction apparatus according to any one of claims 1 to 3, wherein the input unit receives an input of an hourly operation plan of secondary pressure at each payout base or a change thereof. The performance database further stores performance data of hourly discharge amounts of one or more holders on the conduit network;
The first hourly payout amount calculation means further sets the actual condition data of the hourly discharge amount of the holder as the input condition,
The energy load prediction apparatus according to any one of claims 1 to 4, wherein the second hourly payout amount calculation means further sets the hourly operation plan of the holder as the input condition.   6. The energy load prediction apparatus according to claim 5, wherein the input unit receives an input of an hourly operation plan of the holder or a change thereof.

Images