Energy operation plan creation system and energy operation plan creation method

Abstract

To make an energy operation plan at a high speed by using past results of energy operation.SOLUTION: An energy operation planning system 1001 comprises: a demand similarity calculation processing unit 2 which calculates demand similarity in each of pieces of operation result data 12 on the basis of the operation result data 12 and demand prediction data 13; a reference plan selection processing unit 3 which defines time series data of an actual supply value in one piece of operation result data 12 selected on the basis of the demand similarity, as reference plan data 16; a neighbor extraction processing unit 4 which defines time series data of the actual supply value in at least one piece of operation result data 12 selected on the basis of the demand similarity, which does not include the reference plan data 16, as at least one piece of neighbor candidate data 17; and a result information output unit 6 which outputs data having a maximum evaluation value out of the reference plan data and the neighbor candidate data to the outside of the energy operation planning system to provide it to a user as operation plan data 18.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to creating an energy operations plan. [Background technology]

[0002] In recent years, there has been a demand for efficient energy management in order to improve economic efficiency or reduce greenhouse gas emissions. From this perspective, systems that utilize technologies such as the combination of different energy supply equipment, energy trading between multiple regions, and demand response have been established. For the operation of multiple energy sources, technology has been developed to create a plan that minimizes costs by combining supply equipment according to the predicted value of each energy demand. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2009-169930 A [Patent Document 2] JP 2011-130584 A Summary of the Invention [Problem to be solved by the invention]

[0004] The technology disclosed in Patent Document 1 learns from past energy demand records and their fluctuating factors, creates a demand forecasting model, extracts similar days, and predicts future demand. This technology learns from past records, but it is specialized for demand forecasting and cannot be used for operational planning.

[0005] The technology disclosed in Patent Document 2 creates a power generation plan that takes into account prediction errors in renewable energy power generation.

[0006] However, in neither of these techniques is it possible to quickly create an energy management plan because past performance data on energy management is not utilized.

[0007] The present disclosure has been made to solve the above problems, and has an object to quickly create an energy management plan by utilizing past energy management results. [Means for solving the problem]

[0008] The energy management plan creation system of the present disclosure creates operation plan data which is time series data of supply plan values ​​for a future planning period for a plurality of target devices that supply different types of energy. The energy management plan creation system of the present disclosure includes a demand similarity calculation processing unit, a base plan selection processing unit, a neighborhood extraction processing unit, a neighborhood evaluation processing unit, and a result information output unit. The demand similarity calculation processing unit calculates demand similarity which is the similarity between the demand actual value and the demand forecast value in each operation performance data based on a plurality of operation performance data which is time series data of the demand actual value of each energy and the supply actual value by each target device in a plurality of past operation performance periods, and demand forecast data which is time series data of the demand forecast value of each energy in the planning period. The base plan selection processing unit selects the time series data of the supply actual value in one operation performance data selected based on the demand similarity as base plan data, and selects the time series data of the supply actual value in the operation performance data selected based on the demand similarity as the base plan data. If not, adjust the excess or deficiency. The neighborhood extraction processing unit selects time-series data of actual supply values ​​in at least one operation result data that does not include the master plan data, selected based on the demand similarity, as at least one neighborhood candidate data. The neighborhood evaluation processing unit selects at least one neighborhood candidate data based on the demand similarity, and calculates a cost required for the selection as an evaluation value. If not, adjust the excess or deficiency. The cost required for this is calculated as an evaluation value. The result information output unit outputs the data with the highest evaluation value out of the base plan data and the neighboring candidate data to the outside of the energy operation plan creation system to provide it to the user as operation plan data. Effect of the Invention

[0009] According to the energy operation plan creation system disclosed herein, baseline plan data and nearby candidate data are created using time series data of actual supply values ​​in past operation performance data, and the data with the highest evaluation value among these data is used as the operation plan data, thereby enabling the operation plan data to be created quickly. [Brief description of the drawings]

[0010] [Figure 1] 1 is a block diagram of an energy management plan creation device according to a first embodiment. [Diagram 2] 4 is a flowchart of an operation plan creation process according to the first embodiment. [Diagram 3] FIG. 11 is a diagram illustrating an example of device information data. [Figure 4] FIG. 11 is a diagram illustrating an example of operational performance data. [Diagram 5] FIG. 2 is a diagram illustrating an example of demand forecast data. [Figure 6] FIG. 11 is a diagram illustrating an example of operation plan data. [Figure 7] FIG. 11 is a block diagram of an energy management plan creation device according to a second embodiment. [Figure 8] 13 is a flowchart of an operation plan creation process according to the second embodiment. [Figure 9] FIG. 13 is a diagram illustrating an example of a reference plan. [Figure 10] FIG. 13 is a diagram showing an example of a similar plan. [Figure 11] FIG. 11 is a block diagram of an energy management plan creation device according to a third embodiment. [Figure 12] 13 is a flowchart of an operation plan creation process according to the third embodiment. [Figure 13] FIG. 11 is a diagram showing an example of equipment operation data. [Figure 14] FIG. 13 is a diagram illustrating an example of a calculation image of task similarity. [Figure 15] FIG. 11 is a block diagram of an energy management plan creation device according to a fourth embodiment. [Figure 16] 13 is a flowchart of an operation plan creation process according to the fourth embodiment. [Figure 17] It is a block diagram of an energy operation plan creation device according to Embodiment 5. [Figure 18] It is a flowchart of operation plan creation processing according to Embodiment 5. [Figure 19] It is a diagram showing an example of the hardware configuration of an energy operation plan creation device.

Embodiments for Carrying Out the Invention

[0011] <A. Embodiment 1> <A-1. Configuration> FIG. 1 is a block diagram showing an energy operation plan creation system 1001 according to Embodiment 1. The energy operation plan creation system 1001 includes an energy operation plan creation device 101 and a data storage unit 19. The energy operation plan creation device 101 includes an input information acquisition unit 1, a demand similarity calculation processing unit 2, a reference plan selection processing unit 3, a neighborhood extraction processing unit 4, a neighborhood evaluation processing unit 5, and a result information output unit 6.

[0012] The data storage unit 19 stores setting data 14 necessary for the energy operation plan creation device 101 to create an energy operation plan. Hereinafter, the energy operation plan is also referred to as an "operation plan" for short. The setting data 14 includes at least device information data 11, operation result data 12, and demand prediction data 13.

[0013] The input information acquisition unit 1 reads the setting data 14 from the data storage unit 19.

[0014] Based on the setting data 14 read by the input information acquisition unit 1, the demand similarity calculation processing unit 2 calculates demand similarity data 15 between the demand actual data included in the operation result data 12 and the demand prediction data 13.

[0015] Based on the demand similarity data 15, the reference plan selection processing unit 3 selects the operation result data 12 with the highest demand similarity and creates reference plan data 16.

[0016] Based on the demand similarity data 15, the neighborhood extraction processing unit 4 extracts operation performance data with high demand similarity excluding the reference plan, and creates neighborhood candidate data 17.

[0017] Based on the reference plan data 16 and the neighborhood candidate data 17, the neighborhood evaluation processing unit 5 evaluates whether the neighborhood candidate data improves the reference plan data.

[0018] The result information output unit 6 outputs the reference plan data 16 or the neighborhood candidate data 17 with the best evaluation as the operation plan data 18. The operation plan data 18 is stored in the data storage unit 19.

[0019] <A-2. Operation> Figure 2 is a flowchart of the energy operation plan creation process by the energy operation plan creation device 101. Hereinafter, the energy operation plan creation process by the energy operation plan creation device 101 will be described with reference to Figure 2.

[0020] First, in step S101, the input information acquisition unit 1 executes the input information acquisition process. In this process, the input information acquisition unit 1 reads the setting data 14 as the input information. Here, the setting data 14 read by the input information acquisition unit 1 may be the setting data 14 input from the input device 21, or may be the setting data 14 specified by the input device 21 from among the setting data 14 stored on the auxiliary storage device 25 (for example, an electronic file stored in CSV format), or may be the setting data 14 specified by the input device 21 from among the setting data 14 stored in the external storage device 27 (for example, an electronic file stored in CSV format). The input information acquisition unit 1 can read the setting data 14 stored in the external storage device 27 via the network 26. The setting data 14 read by the input information acquisition unit 1 is expanded on the main storage device 24.

[0021] The setting data 14 acquired by the input information acquisition unit 1 in the input information acquisition process includes device information data 11. The device information data 11 is information on devices for which the energy operation plan creation device 101 creates an operation plan. FIG. 3 shows an example of the device information data 11. The device information data 11 includes information such as the name of the device, the type of energy supplied by the device (hereinafter referred to as "supplied energy"), the type of energy consumed by the device (hereinafter referred to as "consumed energy"), and characteristics indicating the amount of energy consumed by the device to supply energy (hereinafter referred to as "consumption characteristics"). In addition, the device information data 11 in FIG. 3 includes maximum and minimum values ​​of the amount of energy supplied by the device.

[0022] The consumption characteristic may be expressed by a fixed value or a polynomial. For example, the consumption characteristic is expressed by a relational expression between supplied energy and consumed energy, such as the following equation (1). Equation (1) expresses the supplied energy as a linear expression of consumed energy.

[0023]

number

[0024] In formula (1), i represents the individual number of the equipment, t represents time, x and y represent energy identification numbers, EneOut represents supplied energy, EneIn represents consumed energy, and a represents a coefficient used in the calculation. x is used to identify the type of supplied energy when there are multiple sources, and y is used to identify the type of consumed energy when there are multiple sources. In formula (1), the consumption characteristics are expressed by a linear expression, but the consumption characteristics may be expressed by a quadratic expression or a polynomial expression, and may include multiple consumed energies. The consumed energy may be fuel or supplied energy generated by other equipment, and the coefficient may differ depending on the heat quantity, quality, state, or temperature.

[0025] The setting data 14 acquired by the input information acquisition unit 1 in the input information acquisition process includes operation performance data 12. The operation performance data 12 includes a value of past energy demand and operation information of the equipment. FIG. 4 shows an example of the operation performance data 12. In FIG. 4, the operation performance data 12 includes a power demand performance value, a heat demand performance value, a steam demand performance value, a gas turbine power supply performance value, a power purchase performance value, and a boiler steam supply performance value. The power demand performance value, the heat demand performance value, and the steam demand performance value are past energy demand performance values ​​recorded for each energy type and for each unit time. The gas turbine power supply performance value and the boiler steam supply performance value are operation information of the equipment recorded for each equipment and for each unit time, and are performance values ​​of the energy supply amount when the equipment is stopped or when the equipment is operating. The operation information of the equipment may include a value of the energy consumption amount during operation, or may include a value of the CO2 emission amount during operation.

[0026] The setting data 14 acquired by the input information acquisition unit 1 in the input information acquisition process includes demand forecast data 13. The demand forecast data 13 includes values ​​of future energy demand. Fig. 5 shows an example of the demand forecast data 13. In Fig. 5, the demand forecast data 13 includes an electricity demand forecast value, a heat demand forecast value, and a steam demand forecast value. These are forecast values ​​of energy demand for each energy type and for each unit time.

[0027] 2, the demand similarity calculation processing unit 2 executes a demand similarity calculation process. The demand similarity calculation processing unit 2 creates demand similarity data 15 by calculating a demand similarity, which is a similarity of an operation result value to a demand forecast value, based on the operation result data 12 and the demand forecast data 13 acquired in step S101. The demand similarity is calculated, for example, by the following formula (2). Formula (2) expresses the demand similarity as the square root of the average of the sum of squares of the difference at each time between the time series data of the demand result value included in the operation result data 12 and the time series data of the demand forecast value included in the demand forecast data 13.

[0028]

number

[0029] In addition, in formula (2), one day is taken as an example. day1 represents the target day of actual demand data, day2 represents the target day of demand forecast, DemSim represents the demand similarity, DemRec represents the actual demand value, DemFore represents the forecasted demand value, and CNT_T_DAY represents the number of demand points included in one day. However, the expression of the demand similarity is not limited to formula (2). For example, the denominator on the right side of formula (2) may be the sum of the absolute value of the actual demand value and the absolute value of the forecasted demand value at each time. Also, instead of calculating the square root on the right side of formula (2), it may be possible to simply use the average of the sum of squares of the differences at each time. Also, the demand similarity may be calculated using a normalized value such that the maximum value of the demand similarity is 1.0.

[0030] The demand score may be 24 points per hour for one day, 48 points per 30 minutes, 48 ​​points per hour or 96 points per 30 minutes for two days, 168 points per week or 336 points per 30 minutes for one week. If the time interval of the time series data is variable, combining different time units such as one hour and four hours, the time series data may be corrected to match the smallest time unit such as one hour and used to calculate the demand similarity.

[0031] For the sake of simplicity, it is assumed in the following description that the operation performance data 12 and the demand forecast data 13 are 24 points of data per day.

[0032] When the operation performance data 12 is data for the past 365 days, 365 demand similarities are calculated for the demand forecast data 13 for the next day. For example, when a user wants to use the operation performance data 12 for the most recent 30 days with a similar climate, 30 demand similarities may be calculated. Also, when the equipment to be planned was updated 100 days ago, 100 demand similarities may be calculated using the operation performance data 12 for the most recent 100 days.

[0033] Next, in step S103 of FIG. 2, the reference plan selection processing unit 3 executes a reference plan selection process. In this process, the reference plan selection processing unit 3 refers to the demand similarity data 15 and selects the time series data of the supply performance value in the operation performance data 12 having the largest demand similarity as the reference plan data 16. The reference plan selection processing unit 3 also calculates the purchase price of energy such as fuel or electricity to be purchased from outside in order to operate the equipment or to satisfy the demand. If the contents of the reference plan indicated in the reference plan data 16 cannot satisfy the forecast demand, the reference plan selection processing unit 3 also calculates the cost required to separately adjust the surplus or shortage. Using these elements, the reference plan selection processing unit 3 calculates the evaluation value of the reference plan by the following formula (3).

[0034]

number

[0035] In formula (3), EvalVal is the evaluation value, EneBuy is the amount of energy purchased from outside, EneTanka is the unit price when purchasing energy from outside, EneAdj is the amount of surplus or shortage compared to demand, and EneAdjTanka is the unit price of energy required to adjust the surplus or shortage. If the energy supply according to the base plan is insufficient compared to demand, EneAdj takes a positive value, and if it is in excess, EneAdj takes a negative value. If there is an excess of energy supply, additional costs may be required to deal with it, but energy may be sold to earn income, so EneAdjTanka can take both positive and negative values. The unit price when purchasing energy, EneTanka, is not the unit price at the time of actual operation, but the unit price for the period covered by the demand forecast.

[0036] Next, in step S104 of Fig. 2, the neighborhood extraction processing unit 4 executes neighborhood extraction processing. In this processing, the neighborhood extraction processing unit 4 refers to the demand similarity data 15 and selects time series data of supply performance values ​​in the operation performance data 12 having a demand similarity close to that of the reference plan data 16 as neighborhood candidate data 17. For example, the neighborhood extraction processing unit 4 sorts the operation performance data 12 in descending order of demand similarity and sequentially selects operation performance data 12 that does not include the reference plan data 16, thereby selecting the operation performance data 12 in descending order of demand similarity and creating the neighborhood candidate data 17. The number of neighborhood candidate data 17 may be one, a number specified by the user, or a number obtained by multiplying the number of operation performance data by a ratio specified by the user.

[0037] Next, in step S105 of Fig. 2, the neighborhood evaluation processing unit 5 executes neighborhood evaluation processing. In this processing, the neighborhood evaluation processing unit 5 calculates an evaluation value EvalVal of the neighborhood candidate indicated by the neighborhood candidate data 17, and if the evaluation value of the neighborhood candidate is higher than the evaluation value of the reference plan, that is, if the neighborhood candidate improves the reference plan, the neighborhood candidate data 17 is set as the operation plan data 18 indicating the supply plan value of each target device in the planning target period. On the other hand, if the evaluation value of the neighborhood candidate is equal to or lower than the evaluation value of the reference plan, the neighborhood evaluation processing unit 5 sets the reference plan data as the operation plan data 18. Note that, if there are multiple neighborhood candidates, the neighborhood evaluation processing unit 5 calculates an evaluation value for each neighborhood candidate, and the neighborhood candidate with the highest evaluation value is compared with the reference plan.

[0038] 2, the result information output unit 6 executes a result information output process. In this process, the result information output unit 6 outputs the operation plan data 18 and stores it in the data storage unit 19. The result information output unit 6 may also display the reference plan on the output device 28 so that the user can check the reference plan.

[0039] FIG. 6 shows an example of the operation plan data 18. The operation plan data 18 in FIG. 6 includes predicted power demand values, predicted heat demand values, predicted steam demand values, gas turbine power supply amounts, power purchase amounts, boiler steam supply amounts, power surpluses / deficits, heat surpluses / deficits, and steam surpluses / deficits.

[0040] <A-3. Effect> The energy operation plan creation system 1001 according to Embodiment 1 includes a demand similarity calculation processing unit 2 that calculates a demand similarity, which is the similarity between the demand actual value and the demand predicted value in each operation actual data 12 based on a plurality of operation actual data 12 and demand prediction data 13; a reference plan selection processing unit 3 that uses the time series data of the supply actual value in one operation actual data 12 selected based on the demand similarity as the reference plan data 16, and calculates the cost required for the reference plan data 16 to satisfy the demand predicted value as the evaluation value; a neighborhood extraction processing unit 4 that uses the time series data of the supply actual value in at least one operation actual data 12 that does not include the reference plan data 16 and is selected based on the demand similarity as at least one neighborhood candidate data 17; a neighborhood evaluation processing unit 5 that calculates the cost required for the neighborhood candidate data 17 to satisfy the demand predicted value as the evaluation value; and a result information output unit 6 that outputs, to the outside of the energy operation plan creation system, the data with the maximum evaluation value among the reference plan data 16 and the neighborhood candidate data 17 as the operation plan data 18 to be provided to the user. Thereby, it is possible to obtain an energy operation plan at high speed without being affected by the complex characteristics of the equipment to be planned.

[0041] <B. Embodiment 2> <B-1. Configuration> FIG. 7 is a block diagram showing an energy operation plan creation system 1002 according to Embodiment 2. The energy operation plan creation system 1002 includes an energy operation plan creation device 102 and a data storage unit 19. The energy operation plan creation device 102 includes a similar plan proposal creation processing unit 41 in addition to the configuration of the energy operation plan creation device 102 according to Embodiment 1. Since most of the energy operation plan creation device 102 is the same as the energy operation plan creation device 101, the description of the overlapping parts with Embodiment 1 will be omitted below.

[0042] The similar plan proposal creation processing unit 41 reads the reference plan data 16, and creates similar plan proposal data 42 by performing at least one of expanding and contracting the operation time and increasing and decreasing the output of the equipment to be planned on the reference plan data 16. The similar plan proposal data 42 includes at least the same items as the operation result data 12 shown in FIG. 4.

[0043] <B-2. Operation> FIG. 8 is a flowchart of an energy operation plan creation method by the energy operation plan creation device 102. The flowchart of FIG. 8 is obtained by adding the similar plan proposal creation process of step S201 between the reference plan selection process of step S103 and the vicinity extraction process of step S104 with respect to the flowchart of FIG. 2.

[0044] In step S201, the similar plan proposal creation processing unit 41 executes the similar plan proposal creation process. The similar plan proposal creation processing unit 41 refers to the reference plan data 16 and creates similar plan proposal data 42 by performing at least one of changing the operation stop state of the equipment included in the reference plan data 16 and changing the energy supply amount.

[0045] Fig. 9 illustrates the baseline plan data 16, and Fig. 10 illustrates similar plan data 42 created by modifying the baseline plan data 16 in Fig. 9. In the similar plan, for example, the start-up time of the equipment is changed by advancing or delaying the time at which the equipment changes from a stopped state to an operating state, and the stop time of the equipment is changed by advancing or delaying the time at which the equipment changes from an operating state to a stopped state.

[0046] For example, whether to bring forward the start time may be determined based on the difference between the actual demand value and the forecast demand value. When the equipment was started at time t in the operation record, the similar plan creation processing unit 41 calculates the demand forecast increase amount=(demand forecast value)-(actual demand value) at time t-1 and time t, respectively, as shown in formula (4), and when the demand forecast increase amount at time t-1 is larger than that at time t, the start time may be brought forward to time t-1.

[0047]

number

[0048] In addition, in the formula (4), DemFore represents a demand forecast value, and DemRec represents an actual demand value.

[0049] The similar plan creation processing unit 41 can increase the energy supply amount in the similar plan when the demand forecast value is greater than the demand actual value, and can decrease the energy supply amount in the similar plan when the demand forecast value is smaller than the demand actual value. When increasing the energy supply amount, the similar plan creation processing unit 41 sets the difference between the demand actual value and the demand forecast value as the increase amount, and if increasing the energy supply amount by that difference would exceed the supply upper limit of the equipment, it only needs to increase the energy supply amount up to the supply upper limit.

[0050] The number of similar plans may be one or a number specified by the user. The similar plan creation processing unit 41 may set the number specified by the user as the upper limit of the number of similar plans and create similar plans only when the condition shown in formula (4) is satisfied.

[0051] Next, in step S104 of FIG. 8, the neighborhood extraction processing unit 4 executes neighborhood extraction processing. In this processing, the neighborhood extraction processing unit 4 selects not only the operation result data 12 with a demand similarity close to the reference plan data 16 but also the similar plan proposal data 42 as the neighborhood candidate data 17.

[0052] Next, in step S105 of FIG. 8, the neighborhood evaluation processing unit 5 executes neighborhood evaluation processing. In this processing, the neighborhood evaluation processing unit 5 calculates the evaluation value EvalVal of the neighborhood candidate. When the evaluation value of the neighborhood candidate is higher than the evaluation value of the reference plan, that is, when the neighborhood candidate improves the reference plan, the neighborhood candidate data 17 is set as the operation plan data 18. The similar plan proposal has the same demand performance value as the reference plan, but the start / stop state of the device or a part of the energy supply amount is different. Therefore, the neighborhood candidate created from the similar plan proposal has a different evaluation value from the reference plan.

[0053] <B-3. Effect> The energy operation plan creation system 1002 according to Embodiment 2 includes a similar plan proposal creation processing unit 41 in addition to the configuration of the energy operation plan creation system 1002 according to Embodiment 1. The similar plan proposal creation processing unit 41 creates the similar plan proposal data 42 by performing at least one of stretching or shortening the operation time of the target device and increasing or decreasing the output of the target device on the reference plan data 16. Also, in the energy operation plan creation system 1002 according to Embodiment 2, the neighborhood extraction processing unit 4 selects the similar plan proposal data 42 as the neighborhood candidate data 17. Therefore, according to the energy operation plan creation system 1002, it is possible to select a highly evaluated operation plan from a wide range of options including the similar plan proposal data 42 obtained by modifying the operation result data 12 in addition to the operation result data 12.

[0054] <C. Embodiment 3> <C-1. Configuration> FIG. 11 is a block diagram showing an energy operation plan creation system 1003 according to Embodiment 3. The energy operation plan creation system 1003 includes an energy operation plan creation device 103 and a data storage unit 19. The energy operation plan creation device 103 includes a work similarity calculation processing unit 62 in addition to the configuration of the energy operation plan creation device 103 according to Embodiment 1. Further, in Embodiment 3, the setting data 14 includes equipment work data 61 in addition to the equipment information data 11, operation result data 12, and demand prediction data 13.

[0055] Since most of the energy operation plan creation device 103 is the same as the energy operation plan creation device 101, the description of the overlapping parts with Embodiment 1 will be omitted below.

[0056] <C-2. Operation> FIG. 12 is a flowchart of an energy operation plan creation method by the energy operation plan creation device 103. The flowchart of FIG. 12 is obtained by adding a similar plan proposal creation process in step S301 between the demand similarity calculation process in step S102 and the reference plan selection process in step S103 with respect to the flowchart of FIG. 2.

[0057] First, in step S101, in the same manner as in Embodiment 1, the input information acquisition unit 1 executes an input information acquisition process. Here, the input information acquisition unit 1 reads the equipment work data 61 in addition to the equipment information data 11, operation result data 12, and demand prediction data 13 described in Embodiment 1.

[0058] FIG. 13 shows an example of the equipment work data 61. The equipment work data 61 includes work information of equipment targeted by the operation plan. The equipment work data 61 includes information such as the name of the equipment, work start date and time, work end date and time, output conditions of the equipment during the work period (hereinafter referred to as "planned conditions"), work content, and worker.

[0059] 12, the task similarity calculation processing unit 62 executes task similarity calculation processing. In this processing, the task similarity calculation processing unit 62 refers to the equipment task data 61, calculates the similarity of the equipment tasks (hereinafter referred to as "task similarity") between the planning period and the period of the operation performance data 12 (hereinafter referred to as "operation performance period"), and creates task similarity data 63.

[0060] Figure 14 shows an example of calculating the task similarity between tasks in the planning period and tasks in the actual operation period. In the example of Figure 14, the upper limit of task similarity is set to 1.0. Also, it is determined whether or not the tasks of the same equipment at the same time are the same, and if they are the same, the task similarity does not decrease.

[0061] 12, the master plan selection processing unit 3 executes master plan selection processing. In this processing, the master plan selection processing unit 3 refers to the operation similarity data 63 in addition to the demand similarity data 15, and selects time-series data of supply performance values ​​in the operation performance data 12 having large demand similarity and operation similarity as the master plan data 16. This processing differs from the first embodiment in that the operation similarity is added to the selection conditions for the master plan.

[0062] Specifically, the master plan selection processing unit 3 calculates an overall similarity from the demand similarity and the operation similarity, and selects the time-series data of the supply performance value in the operation performance data 12 that maximizes the overall similarity as the master plan data 16. For example, the overall similarity is expressed by the following formula (5).

[0063]

number

[0064] In formula (5), α, β, and γ are weighting coefficients for each term, and may be 0.0. Alternatively, the user may specify the reference values ​​of the demand similarity and the work similarity, respectively, and the master plan selection processing unit 3 may calculate the overall similarity using formula (5) for the operation performance data 12 whose demand similarity and work similarity are equal to or greater than the respective reference values.

[0065] After the reference plan selection processing unit 3 selects the reference plan data 16, it calculates the evaluation value of the reference plan in the same manner as in the first embodiment.

[0066] Next, in step S104 of FIG. 12, the neighborhood extraction processing unit 4 executes neighborhood extraction processing. In this processing, the neighborhood extraction processing unit 4 refers to the demand similarity data 15 and the work similarity data 63, and creates neighborhood candidate data 17 from the time series data of the supply performance values in the operation performance data 12 that is close to the reference plan data 16 in terms of the overall similarity. For example, the neighborhood extraction processing unit 4 can sort the operation performance data 12 in descending order of the overall similarity, and select the operation performance data 12 other than the reference plan data 16 in order, so as to select the operation performance data 12 as neighborhood candidates in descending order of the overall similarity.

[0067] Steps S105 and S106 in FIG. 12 are the same as those in the first embodiment.

[0068] <C-3. Effect> The energy operation plan creation system 1003 according to the third embodiment includes a work similarity calculation processing unit 62 in addition to the configuration of the energy operation plan creation systems 1001 and 1002 according to the first embodiment or the second embodiment. The work similarity calculation processing unit 62 calculates the work similarity, which is the similarity of the work of the target device between the planned period and the operation performance period, based on the device work data 61 including information on the work period of the target device and information on the output conditions of the target device during the work period. The reference plan selection processing unit 3 in the third embodiment uses the time series data of the supply performance value in one piece of operation performance data 12 selected based on the demand similarity and the work similarity as the reference plan data 16. The neighborhood extraction processing unit 4 in the third embodiment uses the time series data of the supply performance value in at least one piece of operation performance data 12 that does not include the reference plan data 16 and is selected based on the demand similarity and the work similarity as at least one piece of neighborhood candidate data 17. Therefore, according to the energy operation plan creation system 1003, by considering the work similarity in addition to the demand similarity, an operation plan with a high evaluation can be created.

[0069] <D. Embodiment 4> <D-1. Configuration> FIG. 15 is a block diagram showing an energy operation plan creation system 1004 according to Embodiment 4. The energy operation plan creation system 1004 includes an energy operation plan creation device 104 and a data storage unit 19. The energy operation plan creation device 104 includes a supply amount redistribution processing unit 81 in addition to the configuration of the energy operation plan creation device 101 according to Embodiment 1. The configuration of the energy operation plan creation device 104 other than the supply amount redistribution processing unit 81 is basically the same as that of the energy operation plan creation device 101. Therefore, the description overlapping with that of Embodiment 1 will be omitted below.

[0070] The supply amount redistribution processing unit 81 refers to the reference plan data 16 and recalculates the supply amount of each device according to the predicted demand.

[0071] <D-2. Operation> FIG. 16 is a flowchart showing an energy operation plan creation method processed by the energy operation plan creation device 104 according to Embodiment 4. The flowchart of FIG. 16 adds a supply amount redistribution process of step S401 between the reference plan selection process of step S103 and the vicinity extraction process of step S104 with respect to the flowchart of FIG. 2.

[0072] In step S401, the supply amount redistribution processing unit 81 executes the supply amount redistribution process. In this process, the supply amount redistribution processing unit 81 refers to the demand prediction data 13 and the reference plan data 16 created in step S103, and updates the supply plan value of each device in the reference plan data 16 so that the difference between the total value of the supply plan values of each device and the demand prediction value becomes small. For example, the supply amount redistribution processing unit 81 recalculates the supply amount of each device so as to minimize the difference between the predicted demand and the total value of the supply amounts of each device. Here, the objective function of the recalculation can be expressed as minimizing the square of the difference between the predicted demand of each energy and the total value of the supply amounts at each time, for example, as shown in the following formula (6).

[0073]

Number

[0074] Note that δ represents the weight per energy of the difference between the predicted demand and the total supply volume. Since Equation (6) is expressed as a quadratic equation of continuous variables, for example, using a mathematical programming method such as quadratic programming, the supply volume of each device that minimizes the difference between the predicted demand and the total supply volume of each device can be recalculated.

[0075] The objective function of the redistribution may be not only to minimize the difference between the predicted demand and the total supply volume as expressed in Equation (6), but also to minimize the cost obtained by multiplying the unit price EneAdjTanka of the energy required to adjust the excess or deficiency amount EneAdj to the difference between the predicted demand and the total supply volume.

[0076] Next, in step S104 of FIG. 16, the neighborhood extraction processing unit 4 executes neighborhood extraction processing in the same manner as in the first embodiment. In this processing, the neighborhood extraction processing unit 4 refers to the demand similarity data 15 created in step S102 and selects the operation result data 12 that shows a demand similarity close to the demand similarity of the reference plan data 16. Then, the neighborhood extraction processing unit 4 also executes the same supply volume redistribution processing as in step S401 on the selected operation result data 12, and then saves it as a neighborhood candidate.

[0077] <D-3. Effect> The energy operation plan creation system 1004 according to Embodiment 4 includes a supply amount redistribution processing unit 81 in addition to the configuration of the energy operation plan creation systems 1001-1003 according to any one of Embodiments 1 to 3. The supply amount redistribution processing unit 81 updates the supply plan value of the target device in the reference plan data 16 so that the difference between the predicted demand value and the total supply amount of the target device in the reference plan data 16 becomes small. Also, in Embodiment 4, the neighborhood extraction processing unit 4 updates the supply plan value of the target device in the neighborhood candidate data 17 so that the difference from the total supply amount of the target device in the neighborhood candidate data 17 becomes small. Thereby, according to the energy operation plan creation system 1004, it is possible to quickly obtain an energy operation plan with a small error with respect to the predicted demand.

[0078] <E. Embodiment 5> <E-1. Configuration> FIG. 17 is a block diagram showing an energy operation plan creation system 1005 according to Embodiment 5. The energy operation plan creation system 1005 includes an energy operation plan creation device 105 and a data storage unit 19. The energy operation plan creation device 105 includes a plan model creation processing unit 91 instead of the demand similarity calculation processing unit 2 and an operation plan proposal creation processing unit 93 instead of the reference plan selection processing unit 3, as compared with the energy operation plan creation device 101 according to Embodiment 1, and deletes the neighborhood extraction processing unit 4 and the neighborhood evaluation processing unit 5. Regarding the input information acquisition unit 1 and the result information output unit 6, the configuration of the energy operation plan creation device 105 is the same as that of the energy operation plan creation device 101. Therefore, the description overlapping with Embodiment 1 will be omitted below.

[0079] The plan model creation processing unit 91 reads the setting data 14, learns the operation results of each device from the operation result data 12, and creates plan model data 92. The plan model data 92 is data that determines the supply plan value of each target device with respect to the predicted demand value of each energy.

[0080] The operation plan creation processing unit 93 reads the demand prediction data 13 and the plan model data 92, and creates the supply plan value of each target device defined in the plan model data 92 as the operation plan proposal data 94 corresponding to the demand prediction value based on the demand prediction data 13.

[0081] <E-2. Operation> FIG. 18 is a flowchart of an energy operation plan creation method by the energy operation plan creation device 105 according to Embodiment 5. The flowchart of FIG. 18 is provided with a plan model creation process in step S501 instead of the demand similarity calculation process in step S102, and an operation plan proposal creation process in step S502 instead of the reference plan selection process in step S103, compared to the flowchart of FIG. 2, and eliminates the approximate extraction process in step S104 and the neighborhood evaluation process in step S105.

[0082] In step S501, the plan model creation processing unit 91 executes the plan model creation process. In this process, the plan model creation processing unit 91 creates a plan model, for example, by learning the relationship between the demand performance and the energy supply amount of each device with respect to the operation performance data 12 using machine learning. Specifically, the plan model creation processing unit 91 can create a plan model, for example, by performing supervised learning using the operation performance data 12 as teacher data. Here, the data representing the plan model is referred to as plan model data 92. In other words, the plan model creation processing unit 91 creates the plan model data 92.

[0083] Here, an example in which the plan model creation processing unit 91 learns the relationship between the demand performance and the energy supply amount included in the operation performance data 12 has been given. However, it is not limited to this. When the equipment operation data 61 described in Embodiment 3 is included in the setting data 14, the equipment operation data 61 may also be included in the learning target. The plan model creation process does not need to be executed at the timing of creating the operation plan, and a plan model may be created in advance.

[0084] Next, in step S502, the operation plan proposal creation processing unit 93 executes operation plan proposal creation processing. In this processing, the operation plan proposal creation processing unit 93 creates operation plan proposal data 94 suitable for the predicted demand based on the demand prediction data 13 and the plan model data 92. In this step, by using the plan model data 92 already created in step S501, the time required for creating the operation plan proposal data 94 is reduced.

[0085] Next, in step S106, the result information output unit 6 executes result information output processing. In this processing, the result information output unit 6 outputs the operation plan proposal data 94 and stores it in the data storage unit 19 as operation plan data 18.

[0086] <E-3. Effect> The energy operation plan creation system 1005 according to Embodiment 5 creates operation plan data 18, which is time-series data of supply plan values in the future planned target period of a plurality of target devices that supply different types of energy. The energy operation plan creation system 1005 includes a plan model creation processing unit 91, an operation plan proposal creation processing unit 93, and a result information output unit 6. The plan model creation processing unit 91 learns the operation performance of each target device from a plurality of operation performance data 12, which are time-series data of the demand actual values of each energy and the supply actual values of each target device in a plurality of past operation performance periods, and creates plan model data 92 that determines the supply plan values of each target device for the demand prediction values of each energy. The operation plan proposal creation processing unit 93 creates the operation plan data 18 based on the demand prediction data 13, which is time-series data of the demand prediction values of each energy in the planned target period, and the plan model data 92. The result information output unit 6 outputs the operation plan data 18 to the outside of the energy operation plan creation system 1005 in order to provide it to the user. With the above configuration, the energy operation plan creation system 1005 can obtain an operation plan for the device that matches the demand prediction different from the operation performance. In addition, since the plan model data 92 can be created in advance, only the operation plan proposal creation processing needs to be performed when creating the operation plan data 18, and high-speed operation plan creation is possible.

[0087] <F. Hardware Configuration> FIG. 19 shows an example of the hardware configuration of the energy operation plan creation systems 1001-1005 according to Embodiments 1-5. The energy operation plan creation systems 1001-1005 are composed of an input device 21, a computer 22, an external storage device 27, and an output device 28. The input device 21 is, for example, a keyboard and a mouse. The computer 22 is composed of a CPU 23, a main storage device 24, and an auxiliary storage device 25. The output device 28 is a display device such as a liquid crystal display, and displays the output result from the computer 22.

[0088] The input information acquisition unit 1 is for the user to input data, and is realized by the input device 21. The data storage unit 19 is for storing the setting data 14 and the evaluation result data, and is realized by, for example, any one of the main storage device 24, the auxiliary storage device 25, the network 26, and the external storage device 27, or a combination of a plurality of them.

[0089] The demand similarity calculation processing unit 2, the reference plan selection processing unit 3, the neighborhood extraction processing unit 4, the neighborhood evaluation processing unit 5, the similar plan proposal creation processing unit 41, the operation similarity calculation processing unit 62, the supply amount redistribution processing unit 81, the plan model creation processing unit 91, and the operation plan proposal creation processing unit 93 (hereinafter referred to as "processing units, etc.") are realized by the CPU 23 and the main storage device 24 as internal arithmetic processing of the computer 22. That is, each process in the processing units, etc. is realized by the CPU 23, which is a processor, executing a program developed on the main storage device 24, which is a memory. The CPU 23 reads the data on the main storage device 24, performs each process, and stores the result in the main storage device 24.

[0090] The result information output unit 6 has a function of displaying the operation plan data 18 on the output device 28 so that a user can check the operation plan data 18, for example. In addition, the result information output unit 6 has a function of storing the operation plan data 18 in the auxiliary storage device 25 so as to archive the operation plan data 18. In addition, the result information output unit 6 may store the operation plan data 18 in the external storage device 27 via the network 26.

[0091] The energy management plan creation systems 1001-1005 may be realized by hardware as shown in Fig. 19, or may be realized by software such as a program that embodies each function of the energy management plan creation systems 1001-1005 on a computer, or as a recording medium on which the software is recorded. The recording medium may be any type, such as a hard disk, an optical disk, or a magneto-optical disk.

[0092] It should be noted that the embodiments can be freely combined, and each embodiment can be modified or omitted as appropriate. [Explanation of symbols]

[0093] 1 input information acquisition unit, 2 demand similarity calculation processing unit, 3 base plan selection processing unit, 4 neighborhood extraction processing unit, 5 neighborhood evaluation processing unit, 6 result information output unit, 11 equipment information data, 12 operation performance data, 13 demand forecast data, 14 setting data, 18 operation plan data, 19 data storage unit, 21 input device, 22 computer, 23 CPU, 24 main memory device, 25 auxiliary memory device, 26 network, 27 external memory device, 28 output device, 101-105 energy operation plan creation device, 1001-1005 energy operation plan creation system.

Claims

1. An energy operation plan creation system that creates operation plan data that is time series data of supply plan values ​​in a future planning period for a plurality of target devices that supply different types of energy, comprising: a demand similarity calculation processing unit that calculates a demand similarity, which is a similarity between the demand actual value and the demand forecast value in each of the operation performance data, based on a plurality of operation performance data that is time-series data of a demand actual value of each of the energies and a supply actual value by each of the target devices in a plurality of past operation performance periods, and demand forecast data that is time-series data of a demand forecast value of each of the energies in the planning target period; a master plan selection processing unit that sets time-series data of the actual supply value in one of the operation performance data selected based on the demand similarity as master plan data, and calculates, when the master plan data cannot satisfy the demand forecast value, a cost required to adjust the excess or shortage as an evaluation value; a neighborhood extraction processing unit that selects time-series data of the supply result value in at least one of the operation result data not including the master plan data based on the demand similarity as at least one neighborhood candidate data; a neighborhood evaluation processing unit that, when the neighborhood candidate data does not satisfy the demand forecast value, calculates a cost required to adjust the excess or deficiency as an evaluation value; a result information output unit that outputs the data with the largest evaluation value among the reference plan data and the neighborhood candidate data to an outside of the energy operation plan creation system to provide the data to a user as the operation plan data, Energy operation planning system.

2. A similar plan creation processing unit is further provided for creating similar plan data by performing at least one of extending or shortening an operation time of the target equipment and increasing or decreasing an output of the target equipment with respect to the base plan data, The neighborhood extraction processing unit also selects the similar plan data as the neighborhood candidate data. The energy management plan creation system according to claim 1 .

3. and a task similarity calculation processing unit that calculates a task similarity, which is a similarity of tasks of the target device between the planning target period and the operation performance period, based on device task data including information on a task period of the target device and information on an output condition of the target device during the task period, the master plan selection processing unit sets time-series data of the supply actual value in one of the operation actual data selected based on the demand similarity and the operation similarity as the master plan data, the neighborhood extraction processing unit selects, based on the demand similarity and the operation similarity, time-series data of the supply actual value in at least one of the operation actual data not including the master plan data as at least one of the neighborhood candidate data; The energy management plan creation system according to claim 1 or 2.

4. a supply amount redistribution processing unit that updates the supply plan value of the target equipment in the baseline plan data so as to reduce a difference between the demand forecast value and a total value of the supply amount of the target equipment in the baseline plan data, the neighborhood extraction processing unit updates the supply plan value of the target equipment in the neighborhood candidate data so as to reduce a difference between the supply plan value and a total value of the supply amount of the target equipment in the neighborhood candidate data. The energy management plan creation system according to any one of claims 1 to 3.

5. An energy operation plan creation method by an energy operation plan creation system that creates operation plan data that is time series data of supply plan values ​​in a future planning period for a plurality of target devices that supply different types of energy, comprising: The energy management plan creation system includes a demand similarity calculation processing unit, a master plan selection processing unit, a neighborhood extraction processing unit, a neighborhood evaluation processing unit, and a result information output unit, the demand similarity calculation processing unit calculates a demand similarity, which is a similarity between the demand actual value and the demand forecast value in each of the operation performance data, based on a plurality of operation performance data which are time-series data of demand actual values ​​of each of the energies and supply actual values ​​by each of the target devices in a plurality of past operation performance periods, and demand forecast data which is time-series data of demand forecast values ​​of each of the energies in the planning target period; the master plan selection processing unit sets the time-series data of the supply performance value in one of the operation performance data selected based on the demand similarity as master plan data, the master plan selection processing unit, when the master plan data cannot satisfy the demand forecast value, calculates a cost required to adjust the excess or shortage as an evaluation value; the neighborhood extraction processing unit selects, based on the demand similarity, time-series data of the supply result value in at least one of the operation result data not including the master plan data as at least one neighborhood candidate data; When the neighborhood candidate data does not satisfy the demand forecast value, the neighborhood evaluation processing unit calculates a cost required to adjust the excess or deficiency as an evaluation value; the result information output unit provides the data having the largest evaluation value among the base plan data and the neighborhood candidate data to a user as the operation plan data. How to create an energy operations plan.

Images