Transformer load prediction system and transformer load prediction method

Abstract

To provide a transformer load prediction system and method with which it is possible to find the maximum electric power of a transformer with good accuracy and select a transformer with appropriate capacity as the one to be installed at a site.SOLUTION: In an electric power supply monitoring system, a transformer load prediction device generates a load prediction model in which the number of main switches per capacity that are provided to respective users receiving electric power supply from a transformer is defined as an explanatory variable and the maximum electric power of the transformer is defined as an objective variable, obtains the objective variable by inputting, to the load prediction model, the number of main switches per capacity that are provided to respective users receiving the electric power supply from a prediction object transformer for which the maximum electric power is to be predicted, as the explanatory variable, and outputs the obtained objective variable. Meanwhile, when there is a main switch whose capacity is unknown among the main switches provided to the users, the electric power supply monitoring system identifies the capacity of the main switch whose capacity is unknown, on the basis of the content of contract on electric power supply of the user to whom said main switch is provided.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to a transformer load prediction system and a transformer load prediction method. [Background technology]

[0002] Patent Document 1 discloses a method for estimating maximum load with high accuracy using existing instruments. The method for estimating the load on a transformer is described. The contract capacity and electricity usage of multiple consumers connected to the appliance are acquired, and the contract capacity and electricity usage are used to calculate the A usage estimation function showing the relationship between the contract capacity and the estimated usage is derived from the The load estimate was calculated by estimating the maximum load of the transformer and measured at each ammeter-equipped transformer. The maximum load is obtained by taking the actual load value and calculating the load error, which indicates the error between the estimated load value and the actual load value. The usage error indicates the error between the estimated usage calculated using the usage estimation function and the actual electricity usage. The difference is calculated, and an error function that shows the relationship between the load error and the usage error is derived, and the error function The load estimate is corrected using

[0003] Patent Document 2 describes a method for adjusting the load current of low-voltage customers by utilizing automatic meter readings. Regarding the method for calculating the design capacity of a transformer, which is constructed for the purpose of calculating the capacity of the upper transformer In the above method, low-voltage customers are classified by contract type and their power consumption is recorded on the data. A database will be created to create a standard pattern of electricity usage, and the standard pattern will be compared to the past data. By accumulating data, the increase in electricity usage from the past to the present can be calculated. When determining the capacity of the upper transformer, the contract type and number of low-voltage customers required should be investigated, and the above The contract type load pattern is extracted from the power usage database and the total power for each time period is calculated. The capacity of the pole transformer is calculated by multiplying the increase in electricity usage from the past to the present. do. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2011-205859 A [Patent Document 2] JP 2010-20653 A Summary of the Invention [Problem to be solved by the invention]

[0005] In Patent Documents 1 and 2, the required capacity of a transformer is calculated based on the contract type (contract form) and midnight power However, as the electric power industry becomes more and more divided, the electric power companies may In some cases, it will be impossible to obtain information on the type of sales contract of customers, making it difficult to calculate the capacity of transformers. It is expected that this will happen.

[0006] In addition, traditionally, the calculation of transformer capacity was done based on the number of households and current for each contract type. This is done by calculating the maximum power and simply stacking them together. However, the accuracy of the calculated maximum power is not always sufficient, and transformers with a capacity larger than necessary are often installed on site. It is likely that such measures are often taken.

[0007] The present invention has been made in view of the above background, and provides a method for accurately determining the maximum power of a transformer. The present invention aims to provide a transformer load prediction system and a transformer load prediction method capable of predicting a load of a transformer. The purpose is. [Means for solving the problem]

[0008] In order to achieve the above object, one of the present inventions is a transformer load prediction system, The information processing device includes a processor and a storage device, and receives power from a transformer. The explanatory variables are the number of main switches by capacity installed at each customer receiving the transformer. A load forecasting model is generated, which is a model that uses the maximum power of the device as the objective variable, and the maximum power forecast is The transformer is a target transformer for prediction. The number of main switches by capacity used in the load forecasting model is input as an explanatory variable. and outputting the determined objective variable.

[0009] According to the present invention, it is possible to accurately obtain the maximum power predicted for the transformer to be predicted. Therefore, it is necessary to select a transformer with the capacity required to reliably prevent overloading. It also prevents unnecessarily large capacity transformers from being installed on-site. This allows for efficient operation of transformers (new installations, replacements, etc.).

[0010] Another aspect of the present invention is the transformer load prediction system, There is a main switch installed at a customer receiving power supply, the capacity of which is unknown. If there is a main switch whose capacity is unknown, The capacity is determined based on the contents of the power supply contract.

[0011] According to the present invention, a main power supply system provided in a customer receiving power from a transformer to be predicted is provided. Even if the switch includes a main switch whose capacity is unknown, the predicted capacity is not known for the target transformer. The maximum power that can be obtained can be calculated with high accuracy.

[0012] Another aspect of the present invention is the transformer load prediction system, comprising: Simple regression analysis, multiple regression analysis, DNN (Deep Neural Network), and support vector machine (SVM (Support Vector Machine)) do.

[0013] Another aspect of the present invention is the transformer load prediction system, As explanatory variables, the number of contracts for the supply system (single-phase three-wire 100V) and the total contracted power of metered electricity B The present invention is characterized in that at least one of the total number of contracts with hot water heaters and the number of contracts with hot water heaters is further used.

[0014] By further using these explanatory variables and performing multiple regression analysis, the accuracy of the maximum power prediction can be improved. It can be done.

[0015] Another aspect of the present invention is the transformer load prediction system, The relationship between the number of main switches to be received and the actual measured value of the maximum power of the transformer is stored, and the load prediction model is Based on the model, the relationship between the number of main switches and the predicted maximum power of the transformer is obtained, and and an approximation formula showing a relationship between the number of vehicles and the difference between the predicted value and the actual measured value. The present invention is characterized in that information indicating the difference calculated using the approximation formula is output.

[0016] According to the present invention, the difference between a predicted value and an actual measured value, which is an estimation index of a margin (surplus), can be obtained with high accuracy. Therefore, the predicted maximum power and the obtained margin are used to estimate the target transformer voltage. This allows the transformer capacity required for the device to be determined with precision.

[0017] Other problems and solutions disclosed in this application are described in the section on the preferred embodiment of the invention. and the drawings. Effect of the Invention

[0018] According to the present invention, the maximum power of a transformer can be calculated with high accuracy, and the maximum power of a transformer installed on-site can be calculated with high accuracy. A container with an appropriate capacity can be selected. [Brief description of the drawings]

[0019] [Figure 1] 1 is a diagram showing a schematic configuration of a transformer load prediction system according to an embodiment of the present invention; [Figure 2A] 2 is a diagram illustrating the relationship between a transformer, a smart meter, a main switch, and a load. FIG. [Figure 2B] 1 is a graph illustrating a method for determining the capacity of a transformer. [Diagram 3] FIG. 2 is a diagram showing main functions of a transformer load prediction device. [Figure 4] 2 is an example of a hardware configuration of a transformer load prediction device. [Diagram 5] These are examples of load prediction models (model a to model d). [Figure 6A] FIG. 13 is a diagram showing an example of a simple regression analysis when the capacity of the main switch is 30 A. [Figure 6B] FIG. 13 is a diagram showing an example of a simple regression analysis when the capacity of the main switch is 60 A. [Figure 6C] FIG. 13 is a diagram showing an example of a simple regression analysis when the capacity of the main switch is 100 A. [Figure 6D] FIG. 13 is a diagram showing an example of a simple regression analysis when the capacity of the main switch is 200 A. [Figure 6E] 13 is an example of the results of a simple regression analysis. [Figure 6F] 13 is an example of a result of a multiple regression analysis. [Figure 7] 13 is a graph showing the results of an analysis of the influence of explanatory variables on a response variable (maximum power). [Figure 8A]FIG. 2 is a diagram illustrating an example of consumer information. [Figure 8B] FIG. 11 is a diagram illustrating an example of measurement value information. [Figure 9] FIG. 11 is a system flow diagram illustrating a model learning process. [Figure 10A] 13 is an example of a learning data setting screen. [Figure 10B] This is an example of regional division. [Figure 11] FIG. 11 is a system flow diagram illustrating a load prediction process. [Figure 12] 13 is an example of a result display screen. [Figure 13A] 1 is a graph showing the relationship between the number of contracts and maximum power consumption based on actual measurements. [Figure 13B] 1 is a graph showing the results of predicting the relationship between the number of contracts and maximum demand using a conventional method. [Figure 13C] 13 is a graph showing the results of predicting the relationship between the number of contracts and maximum demand using model a. [Figure 13D] 13 is a graph showing the results of predicting the relationship between the number of contracts and maximum demand using model b. [Figure 13E] 13 is a graph showing the results of predicting the relationship between the number of contracts and maximum demand using model c. [Figure 13F] 13 is a graph showing the results of predicting the relationship between the number of contracts and maximum demand using model d. [Figure 13G] This is a table showing the coefficient of determination R2 for easy comparison. [Figure 13H] 13 is a graph showing the results of predicting the relationship between the number of contracts and maximum demand using model b. [Figure 13I] This is a table showing the coefficients and intercepts of each explanatory variable in model b. [Figure 13J] 13 is a graph showing the results of predicting the relationship between the number of contracts and maximum demand using model d. [Figure 13K] This is a table showing the coefficients and intercepts of each explanatory variable in model d. [Figure 14A] This is a graph plotting the difference between the predicted values ​​and the actual values ​​for each number of contracts for model b, and shows the upper limit of the interval in which 95% of the predicted values ​​fall. [Figure 14B] This is a graph plotting the difference between predicted values ​​and actual values ​​for each number of contracts for model d, and shows the upper limit of the interval in which 95% of the predicted values ​​fall. [Figure 14C] This is a graph plotting the difference between the predicted values ​​and the actual values ​​for each number of contracts for model b, and shows the upper limit of the interval in which 99.7% of the predicted values ​​fall. [Figure 14D] This is a graph plotting the difference between the predicted values ​​and the actual values ​​for each number of contracts for model D, and also shows the upper limit of the interval in which 99.7% of the predicted values ​​fall. [Figure 15A] 14B is a graph showing the results of approximating, using an approximation formula, the upper limit of an interval in which 95% of predicted values ​​fall, based on FIG. 14A. [Figure 15B] 15B is a table showing the parameters of each approximation formula in FIG. 15A, the ratio of the number of times that the predicted value exceeds the actual measured value, and the average value of the above differences. [Figure 15C] 14C is a graph showing the results of approximating, using an approximation formula, the upper limit value of an interval in which 95% of the predicted values ​​fall, based on FIG. 14B. [Figure 15D] 15C , the ratio of the number of times the predicted value exceeds the actual measured value, and the average value of the above differences. [Figure 15E] 14D is a graph showing the results of approximating, using an approximation formula, the upper limit value of the interval in which 99.7% of the predicted values ​​fall, based on FIG. 14C. [Figure 15F] 15E is a table showing the parameters of each approximation formula in FIG. 15E, the ratio of the number of times that the predicted value exceeds the actual measured value, and the average value of the above differences. [Figure 15G] 14E is a graph showing the results of approximating, using an approximation formula, the upper limit value of the interval in which 99.7% of the predicted values ​​fall, based on FIG. 14D. [Figure 15H] 15G, the ratio of the number of times that the predicted value exceeds the actual measured value, and the average value of the above differences. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0020] Hereinafter, an embodiment will be described with reference to the drawings. Similar configurations may be denoted by common reference symbols and descriptions thereof may be omitted.

[0021] FIG. 1 shows an information processing system (hereinafter, referred to as a “transformer load prediction system”) according to an embodiment of the present invention. As shown in the figure, the transformer load prediction system The system 1 is a power information management system installed in a system center 3 of an electric power company or other electric utility. A load prediction device 100 and a load prediction device 4 for transformer A number of smart meters SM are provided at consumers 2 that receive power via a device; The system center 3 may be, for example, a control center for a smart grid or a power The consumer 2 is, for example, a general household, a public facility, These include office buildings, various public facilities, factories, and infrastructure facilities.

[0022] The power information management device 4 is configured using an information processing device (computer) and and a smart meter SM installed in the The communication network 5 is, for example, a WAN (Wide Area Network). power supply, concentrator, 920Hz band communication network, PLC (Power Line Communication ) and other communication infrastructure.

[0023] The smart meter SM measures the electricity supplied from the power transmission and distribution system to the consumer 2 in which it is installed. The measured power value is sent to the power information management device 4 via the communication network 5 at regular intervals (every 30 minutes, etc.). The power information management device 4 receives various measurement values ​​(power Based on the values ​​of the voltage, current, active power, reactive power, power factor, temperature, etc., the power supply Provides information to monitor for any abnormalities in supply.

[0024] The load prediction device 100 is configured using an information processing device and is provided in a system center 3. via a communication means such as a LAN (Local Area Network) or a communication network 5, The load prediction device 100 is communicably connected to the power information management device 4. The measurement values ​​obtained from the smart meter SM and information on the consumer 2 are provided by the device 4. Using this information, the maximum power (maximum load) of the transformer supplying power to each consumer 2 is predicted. In addition, the load prediction device 100 estimates the margin ( The margin is calculated and the calculated margin is presented to the user together with the maximum power (maximum load). Based on the maximum power and margin presented by the load prediction device 100, a transformer to be installed at the site is Therefore, it is necessary to efficiently select and decide on a transformer with an appropriate capacity that is not excessive and prevents overloading. It can be determined.

[0025] FIG. 2A shows a transformer 7 (such as a pole transformer or an underground transformer) that supplies power to a consumer 2 and a This is a diagram illustrating the relationship between devices installed in a house 2. As shown in the figure, a transformer 7 is The power transmitted from the power distribution system 8 is transformed and one of the subordinate (transformer 7 bank) The transformer 7 supplies electricity to the above-mentioned consumers 2. A smart meter SM is connected via a power supply (such as a three-phase three-wire system), and the smart meter SM and the load L The main switch MS (breaker) is provided between the load and the load. An appropriate capacity should be selected to prevent overloading and to avoid overloading. For example, From among multiple types of transformers 7 (5kW, 10kW, 20kW, 30kW, 50kW, etc.) , the predicted maximum power (maximum load) plus a margin is equal to or greater than the smallest capacity. The capacity of the main switch MS (30A, 40A, 50A, 60A, 75A , 100A, 125A, 150A, etc.) are between the consumer 2 and the power supplier such as the power company. Determined by the power supply contract or expected maximum load.

[0026] FIG. 2B is a diagram for explaining a method for determining the capacity of a transformer. Here, the capacity of a transformer shown in FIG. 2A is used as an example. The method of determining the capacity of the transformer 7A shown in FIG. The vertical axis is the power. The solid lines P, Q, and R are the power supplies from the transformer 7A. The power consumption of the three loads L (P, Q, R) (measured by the smart meter SM) is The dashed line G represents the total power consumption of the three loads L (P, Q, R). During the period shown in the figure, the total value around 12:00 was the maximum power (7.8 kW). The capacity of the transformer 7A must be 7.8kW or more. For example, , a capacity of 10 kW is selected.

[0027] FIG. 3 is a block diagram showing main functions of the load prediction device 100. As described above, the load prediction device 100 includes a storage unit 110, a consumer information management unit 120, a measurement value acquisition unit 1 30, learning data generation unit 135, model learning unit 140, prediction target transformer information acquisition unit 145 The system includes the functions of a load prediction unit 150, a margin calculation unit 155, and a result presentation unit 160.

[0028] Among the above functions, the storage unit 110 stores consumer information 111, measurement value information 112, learning data 113, a load forecasting model 114, forecast target transformer information 115, a load forecasting result 116, and Each piece of information (data) of the margin calculation result 117 is stored.

[0029] The consumer information 111 stores information about each consumer 2 (hereinafter, referred to as "consumer information"). The consumer information 111 is obtained from the user via the user interface, and also from the power Various systems operated by electric power companies (power distribution control systems and customer management systems) The consumer information 111 is obtained from the latest consumer information and past consumer information of each consumer 2. The customer information includes the customer information from the past. Specific examples of the customer information will be described later.

[0030] The measurement value information 112 is a predetermined time measured by the smart meter SM for each consumer 2. The power value for each period (the power information management device 4 transmits the power value from the smart meter SM of each consumer 2 to the communication network) Time series data of measured values ​​acquired via Network 5 (data in which time information is associated with measured values) Hereinafter, this will be referred to as "measurement value information." The measurement value information 112 includes the most recently acquired In the measurement value information 112, each measurement value information includes information obtained by the measurement value information 112 and information obtained in the past. The data is managed in association with the identifier of the smart meter SM from which it was acquired (the meter ID, which will be described later). The substance of the measurement information 112 is, for example, a file or data managed by a file system. Base tables, network storage storage areas (e.g. URLs (Uniform Resource Locators) The data is managed in a storage area specified by a data store Locator.

[0031] The learning data 113 is learning data (teacher data) used for learning a load prediction model 114 described later. The learning data 113 includes the measurement value information 112 or the measurement value information 112. It is generated from information obtained by interpolating missing information. Details of training data 113 More on this later.

[0032] The load forecast model 114 is a model for predicting loads of the loads that are under its control (powered (present in the bank) The transformer for which the maximum power consumption of consumer 2 is predicted (hereinafter, “prediction Information on the transformer to be measured (hereinafter referred to as "transformer to be predicted") (hereinafter referred to as "transformer information to be predicted") A machine learning model that outputs the maximum power as the objective variable when the following is input as an explanatory variable: The load forecast model 114 may be implemented as a polynomial model including adjustable parameters. The load prediction model 114 is an equation, a determinant, a mathematical expression, a vector, etc. The details of the load prediction model 114 will be described later. .

[0033] The prediction target transformer information 115 is input as an explanatory variable to the load prediction model 114. The prediction target transformer information includes, for example, the load prediction device 100 based on the customer information. The information is automatically generated based on the information 111 and the measurement value information 112, or the user inputs the information. Input via the interface.

[0034] The load forecast result 116 is obtained by inputting the above explanatory variables into the load forecast model 114. Including the objective variable output by the load prediction model 114 (hereinafter referred to as the "load prediction result"). .

[0035] The margin calculation result 117 includes the margin calculated by the margin calculation unit 155 .

[0036] Among the functions shown in FIG. 3, the customer information management unit 120 manages the customer information 111 ( The customer information management unit 120 performs, for example, the following operations: acquiring, editing, registering, and deleting customer information. A user interface is provided for managing the above-mentioned information.

[0037] The measurement value acquisition unit 130 receives the power information sent from the power information management device 4 at a predetermined time interval (every 30 minutes, etc.). The measurement value information received is acquired and managed as measurement value information 112.

[0038] The learning data generation unit 135 generates learning data based on the consumer information 111 and the measurement value information 112. Generate 113 (teaching data).

[0039] The prediction target transformer information acquisition unit 145 acquires prediction target transformer information.

[0040] The load prediction unit 150 inputs the prediction target transformer information to the load prediction model 114 as an explanatory variable. The future maximum power of the transformer to be predicted is calculated by inputting the load prediction unit 150. The maximum power is stored in the storage unit 110 as the load prediction result 116 .

[0041] The margin calculation unit 155 calculates the maximum power consumption predicted value obtained by the load prediction unit 150 and the maximum power consumption The margin is calculated based on the actual measured value (measurement value). 10 stores the result as a margin calculation result 117. The method of calculating the margin will be described later in detail.

[0042] The result presentation unit 160 presents the load prediction result 116 and the margin calculation result 117 through a user interface. The information is presented to the user via the

[0043] FIG. 4 shows an example of a hardware configuration of an information processing device used to realize the load prediction device 100. The illustrated information processing device 10 includes a processor 11, a main memory device 12, an auxiliary memory device 1 3, an input device 14, an output device 15, and a communication device 16. For example, personal computers, office computers, various server devices, The information processing device 10 may be, in whole or in part, a cloud system. Virtual servers provided by the like, using virtualization technology and process space isolation technology, etc. The load prediction device may be realized by using the virtual information processing resources provided. The information processing device 100 may be realized using a plurality of information processing devices 10 connected so as to be capable of communicating with each other.

[0044] In the figure, the processor 11 is, for example, a CPU (Central Processing Unit), PU(Micro Processing Unit), GPU(Graphics Processing Unit), FPGA(Fie ld Programmable Gate Array), ASIC (Application Specific Integrated Circuit) ), AI (Artificial Intelligence) chips, etc.

[0045] The main memory device 12 is a device for storing programs and data, and is, for example, a ROM (Read Only Memory). Only Memory), RAM (Random Access Memory), Non-volatile Memory (NVRAM (Non Volatile RAM), etc.

[0046] The auxiliary storage device 13 is, for example, a solid state drive (SSD) or a hard disk drive. Optical storage devices (CD (Compact Disc), DVD (Digital Versatile Disc), etc.), Reading / writing of storage systems, IC cards, SD cards, optical recording media, etc. The auxiliary storage device 13 includes a reading device for a recording medium, a storage area of ​​a cloud server, etc. The auxiliary storage device 1 can read programs and data via the communication device 16. The programs and data stored (memorized) in 3 are loaded into the main memory device 12 as needed.

[0047] The input device 14 is an interface that accepts input from the outside, and is, for example, a keyboard. Keyboards, mice, touch panels, card readers, pen-input tablets, voice input devices etc.

[0048] The output device 15 is an interface that outputs various information such as the process and results of processing. The output device 15 is, for example, a display device (LCD (Liquid Crystal Display)) that visualizes the various information described above. al Display, graphic card, etc.), devices that convert the above various information into voice (voice output devices, A device that converts the above information into text (such as a printer). For example, the information processing device 10 inputs and outputs information to and from other devices via the communication device 16. This may also be configured.

[0049] The input device 14 and the output device 15 receive and present information to and from the user. Configure the user interface.

[0050] The communication device 16 communicates with other devices via a communication infrastructure such as a communication network 5 ( A device that realizes wired or wireless communication, such as a NIC (Network Interface Card). rd), a wireless communication module, a USB module, etc.

[0051] The information processing device 10 includes, for example, an operating system, a file system, a database, MS (DataBase Management System) (relational database, NoSQL, etc.), KVS (Key-Value Store) or the like may be introduced.

[0052] The functions of the load prediction device 100 are implemented by the processor 1 of each of the information processing devices 10. 1 reads and executes the programs stored in the respective main memory devices 12. Alternatively, the hardware constituting the load prediction device 100 (FPGA, ASIC, AI The load prediction device 100 is realized by the functions of the load prediction device itself (such as a chip). (Data) is, for example, a table in a database or a file managed by a file system. Store it as.

[0053] =Load forecast model= In this embodiment, in a case where the capacity of the main switch MS installed in the consumer 2 cannot be acquired, Taking this into consideration, the load forecasting model is expressed as follows: [Number 1] Load forecast model = Maximum power (customer with unknown main switch capacity) + Maximum power (customers with known main switch capacity) ...Formula 1

[0054] Here, the maximum power in the first term of the above formula is calculated using the conventional formula for calculating maximum power (hereinafter referred to as the "conventional formula"). ) can be used, and the main switch capacity can be calculated based on the actual (known) contract details (contract By replacing the variables with the corresponding variables (such as the type of analysis), it is also possible to perform simple and multiple regression analysis. On the other hand, the second term of the above formula is not applicable because the contract details (contract type, etc.) are unknown. cannot be used and must be determined by either simple regression analysis or multiple regression analysis.

[0055] Therefore, in this embodiment, the four models shown in FIG. 5, which are combinations of the first and second terms of the above formula, are used. We investigated each of the models (model a to model d).

[0056] <Conventional> An example of the conventional formula used for models a and c is shown in the following formula. [Number 2] P1 = a + b + c + d ··· Formula 2 In the above formula, P1 is the expected peak load, a is the flat-rate equivalent (Σ(kW)), and b is the metered lighting equivalent A. (Σ(current value × number of households × voltage)), c is the TOU equivalent (Σ(current value × 1.5 × number of households × voltage)), d is This is the equivalent of metered electricity B (Σ(kW×0.7)). Σ means the sum for each customer (contract). It is.

[0057] When applying the above formula, for example, the current value per household of customer 2 of the metered electricity A is calculated based on the past Based on the measured values, the estimated value is based on the number of households (in increments of 1) that use metered electricity A in the same bank. In the example, for time-of-day lighting contracts, the assumed current value is multiplied by 1.5, and for metered lighting B, The contracted power (KVA) is multiplied by 0.7.

[0058] <Simple regression analysis> For the first term of model a or the simple regression analysis of model b, the main switch MS capacity of each unit is The study was carried out by calculating the slope and intercept for each capacity of the main switch MS using the maximum power of the number of switches as an explanatory variable. I gave evidence.

[0059] 6A to 6D show an example of a simple regression analysis. When the capacity of S is 30A, 60A, 100A, 200A, the number of units (number of contracts) and the number of units 13 is a graph showing an example of the results of a simple regression analysis of the relationship between the maximum power and the

[0060] FIG. 6E shows the capacity of the main switch MS of 30A, 40A, 50A, 60A, and 75A. By performing a simple regression analysis for A, 100A, 125A, 150A, and 200A The explanatory variables used to calculate the slope and intercept in Table 2 are shown in Figure 5. A different one was used from that used in Figures 5A to 5D.

[0061] An example of the first term of model a (or model b) configured based on FIG. 6E is shown in the following equation. [Number 3] The first term of model a (or model b) =((1.299×n+2.120)×n 30 +(1.279×n+2.029)×n 40 +··· +(4.928×n+5.501)×n 150 +(6.544×n+7.199)×n 200 ) / n ...Formula 3 In the above formula, n is the total number of contracts in the bank of the transformer to be forecasted, and n x is the main switch The number of units by capacity (x is capacity [A]).

[0062] <Multiple regression analysis> For the second term of model c or the multiple regression analysis of model d, the explanatory variables were (registered trademark) into the statistical model of multiple regression analysis to obtain the coefficient (constant) of the number of units by main switch capacity. Verification was performed by determining the

[0063] FIG. 6F shows an example of values ​​output by the above statistical model of multiple regression analysis. The estimated value (estimate) of the coefficient (including the constant part) of the number of units by capacity, as well as the standard error (stderr) , t-value, p-value (p>|t|), confidence interval ([0.025 0.975]), coefficient of determination (R 2 ) .

[0064] An example of the second term of model c or model d configured based on FIG. 6F is shown in the following equation. [Number 4] Maximum power [kW]=0.4042n 30 + 0.9847n 40 + 1.7370n 50 +2.0604n 60 + 6.5222n 200 + 5.1914 ...Formula 4 In the above formula, n x is the number of main switchgear of each capacity (x is the capacity [A]).

[0065] <Verification of adding explanatory variables> In order to improve the accuracy of multiple regression analysis, we investigated the addition of explanatory variables (feature values). Specifically, it is a Python (registered trademark) data analysis library called "XGboost " was used to rank the degree of influence of the explanatory variables on the objective variable (maximum power).

[0066] The result of the ranking is shown in FIG. 7. In this embodiment, the top three explanations in the figure are Variables, i.e., "Number of contracts for supply method (single-phase three-wire 100V)", "Total contract power for metered electricity B" A multiple regression analysis was performed by adding the following variables: "Number of contracts with water heaters" and "Number of contracts with water heaters" to confirm their impact on accuracy.

[0067] =Learning Data= The learning data (teaching data) used for learning the load forecasting model 114 is the data of the power information management system. The measurement values ​​(3) sent from the smart meters SM of each consumer 2 are provided by the management device 4. The measured values ​​were randomly extracted and used as explanatory variables. In the case of losses, the average value of contracts with the same attributes was used for interpolation. The power values ​​(30-minute values) sent from the UPS were converted to instantaneous values ​​(1-minute values) using the following formula. [Number 5] Instantaneous value (winter peak) = 30-minute value × 1.425 + 0.149 (kW) Formula 5 [Number 6] Instantaneous value (summer peak) = 30-minute value × 1.347 + 0.050 (kW) Formula 6

[0068] Using the information obtained above, calculate the sum of instantaneous values ​​for each hour / transformer for a specified period. The maximum value of the Successful.

[0069] =Customer information= An example of the customer information 111 is shown in FIG. 8A. As shown in the figure, the customer information 11 1 is the main switch capacity 111a, the meter ID 111b, the establishment code 111c, and the establishment name 11 1d, transformer pole number 111e, service pole number f, contract type code 111g, contract classification 1 11h, contract power (sub-B) 111i, supply method code 111j, supply method 111k, electricity Water heater availability 111L, electric water heater capacity (heat pump) 111m, number of electric water heaters (heat Pump) 111n, Hot water heater capacity 111o, Number of hot water heaters 111p, Renewable energy availability 111q, Multiple records including information on each item such as latitude 111r, longitude 111s, and location of use 111t One record in the customer information 111 is composed of one customer (one It supports one measuring instrument (smart meter) or one main switch (MS).

[0070] Among these, the main switch capacity 111a is the capacity of the main switch MS installed in the customer 2. The main switch capacity 111a is information indicating the amount of the main switch capacity, for example, when a new switch is installed, replaced, or inspected. The capacity of the main switch 111a is acquired when the customer 2 is supplied with power. The capacity estimated from the material of the supplying drop line may be used.

[0071] The meter ID 111b is the power meter (smart meter S The instrument ID, which is the identifier for the meter (M), is stored.

[0072] The business establishment code 111c is the electric power company that supplies power to the customer 2. The establishment code, which is an identifier of the business establishment of the business operator, is stored. The name of the business is stored.

[0073] The transformer pole number 111e is an identifier of the transformer pole supplying power to the customer 2. The transformer pole number f is stored. The service pole number f is stored. The number of the lead-in pole is stored.

[0074] The contract type code 111g indicates the contract type that the consumer 2 has concluded with a power supply company such as a power company. The contract type code, which is information indicating the classification of the contract, is stored in the contract classification 111h. Contract classification, which is information indicating the type of classified contract, is stored. Contract Power (Sub-B) 111 In i, the contracted power when the contract classification is "metered lighting B" is stored.

[0075] The supply method code 111j is a code indicating the method of supplying power to the consumer 2. The supply method code (1: single-phase two-wire, 2: single-phase three-wire, etc.) is stored. The information indicating the supply method (single-phase two-wire system or single-phase three-wire system, etc.) is stored in this register.

[0076] The electric water heater presence / absence 111l indicates whether or not an electric water heater is installed in the consumer 2. The electric water heater capacity (heat pump) 111m stores the information on the demand Information indicating the capacity of the heat pump type electric water heater installed in house 2 is stored. The number of heat pumps installed in the customer 2 is included in the number of heat pumps installed in the customer 2. The number of pump-type electric water heaters is stored.

[0077] The water heater capacity 111o stores information indicating the capacity of the water heater installed in the consumer 2. The number of hot water heaters 111p stores the number of hot water heaters installed in the customer 2. will be done.

[0078] The renewable energy availability 111q section contains the renewable energy generator at the consumer 2. Information indicating the presence or absence is stored.

[0079] The latitude 111r and the longitude 111s are the latitude and longitude of the service pole to which the service line to the customer 2 is connected. The location of use 111t is the pole to which the service line to the customer 2 is connected. Information indicating the location of is stored.

[0080] FIG. 8B shows an example of the measurement value information 112. As shown in the figure, the measurement value information 112 includes the following: A plurality of records having items such as an instrument ID 1121, a measurement date and time 1122, and a power value 1123. One record of the measurement value information 112 is sent from a certain smart meter SM. The measurement value corresponds to a measurement value at a certain measurement date and time.

[0081] Among the above items, the meter ID 1121 stores the meter ID of the smart meter SM. The measurement date and time 1122 stores the date and time when the measurement value was measured. The power value (30-minute value) measured by the smart meter SM is stored in the field.

[0082] =About margin= As mentioned above, in order to avoid overloading of the transformer 7, the load prediction model 114 predicts In this embodiment, it is necessary to set an appropriate margin for the predicted value. The upper limit of the interval in which 95% of the measured values ​​fall (average value + standard deviation σ × 2) for the difference between the Calculate the upper limit of the interval in which 99.7% of the measured values ​​fall (average value + standard deviation σ × 3), and Generate an approximation formula for the upper limit (the relationship between the number of contracts (number of customers 2) for each transformer 7 and the above difference) The results of a specific study on the above approximation will be described later.

[0083] = Model Training = FIG. 9 shows a load prediction device 100 in which the learning data generating unit 135 and the model learning unit 140 The generation of learning data and the training of the load forecasting model 114 using the generated learning data (conventional (Including learning about coefficients and constants in linear regression analysis and learning about coefficients and constants in simple regression analysis and multiple regression analysis.) A flow chart for explaining the process (hereinafter referred to as "model learning process S900") performed when It is a chart.

[0084] The timing for executing the model learning process S900 is not necessarily limited. For example, In response to the generation of new learning data 113 and the user interface, the load prediction device 100 When an explicit execution command is received from the user via the interface, the model learning The model learning process S900 is described below with reference to FIG. do.

[0085] First, the learning data generation unit 135 receives from the user the range of measurement values ​​( The designation (extraction conditions) (regional classification, target period, transformer (bank), etc.) was accepted and accepted. Generate training data based on the measurement values ​​corresponding to the range, and use the generated training data as training data 1 Store this as 13 (S911).

[0086] FIG. 10A shows how the learning data generating unit 135 displays a user interface when accepting the above designation. A screen presented to the user via the browser (hereinafter referred to as a "learning data setting screen 1000"). As shown in the figure, the learning data setting screen 1000 shown in the example includes a region division 1010, 1020, 1030, and a setting button 1050.

[0087] The area classification specification field 1010 contains information specifying the area in which the transformer is installed. Set the region division. The region division specification field 1010 is in the form of a pull-down menu. In addition to “ALL” which specifies all regions, the user can select one or more regions as shown in FIG. 10B. A region division can be specified.

[0088] The period of measurement is specified in the target period specification field 1010. is in the form of a pull-down menu, and users can select "ALL" to specify the entire period, One or more periods can be specified.

[0089] In the transformer (bank) specification field 1030, information for specifying the transformer (bank) (in this example, The transformer (pole number) can be specified. The transformer (bank) specification field 1030 is The menu is a drop-down menu, and the user can select "AL In addition to "L", specify one or more transformer (bank) identifiers (transformer pole number, etc.) can be done.

[0090] When the user operates the setting button 1050, the learning data generating unit 135 The measurement value corresponding to the designated content is acquired (extracted) from the measurement value information 112 and used as learning data. The generated learning data is stored in the storage unit 110 as learning data 113. It should be noted that the specification field shown in the figure is merely an example, and other conditions may be specified.

[0091] Returning to FIG. 9, the model learning unit 140 then creates a load prediction model based on the learning data 113. The model learning unit 140 learns the load forecast rule 114 (S912). The prediction accuracy of the measurement model 114 may be verified. In that case, for example, The learning data 113 is classified into learning data and verification data in advance, and a load prediction model The learning of 114 is done using learning data, and the load prediction model 114 is verified. When carrying out the verification, verification data is used.

[0092] FIG. 11 shows the results of the load predicting device 100 predicting the maximum power and calculating the margin. A flow chart for explaining the process presented to the user (hereinafter referred to as "load prediction process S1100"). The load prediction process S1100 is performed at a predetermined timing (periodically, The execution time is based on the date and time when the program is downloaded, or when an execution instruction is received from the user. The load prediction process S1100 will now be described.

[0093] First, the prediction target transformer information acquisition unit 145 acquires the following from the customer information 111 and the measurement value information 112: , and obtains prediction target transformer information 115, which is an explanatory variable to be input to the load prediction model 114. (S1111). The explanatory variables are the information obtained from the customers (the new transformer construction shown in the figure). The new transformer installation information 118 may include, for example, information about an apartment complex or When installing a new transformer for a new apartment or condominium, the expected load (main switch capacity) In addition, the new transformer installation information 118 includes, for example, For example, when supplying electricity from an existing transformer to a new home or store, the information of the new customer must be collected. In this case, the required capacity of the transformer, including existing contracts, is registered as an explanatory variable. The load forecast results are used, for example, to determine whether or not the capacity of existing transformers needs to be increased. Used.

[0094] Next, the load prediction unit 150 inputs the prediction target transformer information 115 into the load prediction model 114. The maximum power is obtained as the output (S1112).

[0095] Next, the tolerance calculation unit 155 calculates the tolerance (S1113).

[0096] Next, the result presentation unit 160 presents the maximum power calculated in S1112 and the margin calculated in S1113. The screen on which the results are displayed (hereinafter referred to as the "result display screen 1200") is displayed on the output device 15 (display device). The result is output (displayed) to the display device (S1114).

[0097] An example of the result display screen 1200 is shown in FIG. A display field for the pole number 1210, a display field for the maximum power (predicted value) 1220, and a display field for the margin 1 230, and a display field 1240 for the recommended transformer capacity.

[0098] In the display field 1210 of the transformer pole number, the transformer pole number of the transformer to be predicted (the transformer pole number of the transformer to be predicted) is displayed. In the display field 1220 of the maximum power (predicted value), The predicted value of the maximum power of the transformer to be predicted, predicted by the load prediction unit 150, is displayed. The tolerance calculated by the tolerance calculation unit 155 is displayed in a display field 1230. The capacity display field 1240 displays the maximum power (estimated value) of the selectable transformer capacities. The predicted maximum power value displayed in 1220 is matched with the predicted margin displayed in the margin display field 1230. The transformer capacity that is equal to or greater than the value to which the measured value is added and that is closest to the value is automatically selected and displayed. In addition, instead of automatic selection, candidate transformer capacities are presented and the user is asked to manually select the transformer capacity. Alternatively, the user may manually select or decide the desired item.

[0099] As described above in detail, according to the transformer load prediction system 1 of the present embodiment, The maximum power predicted for the transformer to be predicted is obtained with high accuracy using the measurement model 114. In addition, the tolerance calculation unit 155 can calculate the tolerance with high accuracy. Based on the predicted maximum power and the obtained margin, the transformer voltage required for the transformer to be predicted is calculated. This allows the capacity of the device to be determined with high accuracy. This makes it possible to accurately and appropriately select a transformer with a suitable capacity. This can prevent a large number of transformers from being installed on-site, and can make transformer operation (new installation, replacement, etc.) more efficient. This makes it possible to do so efficiently.

[0100] =Verification results= Next, the results of various verifications performed on the transformer load prediction system 1 described above will be described. I will explain it below.

[0101] <Comparative verification of models> For each of the four load forecasting models 114 (models a to d) shown in Table 1, Based on the switch capacity and contract details (contract type, etc.) and the power value obtained from the smart meter SM Four load forecasting models 114 are generated by learning using the learning data based on the above. The results were compared.

[0102] 13A and 13B are both comparative examples, and among them, FIG. 13A shows the measured values ​​(actual values ) based on the conventional method, and Fig. 13B is a graph showing the relationship between the number of contracts and maximum power consumption. 13B is a graph showing the relationship between the maximum power and the coefficient of determination R 2 Also show The coefficient of determination R 2 The maximum value of is 1.0. The coefficient of determination R 2 The closer the value of is to 1.0, the better the model is. This indicates that the accuracy is good (the predicted value closely represents the actual measured value).

[0103] 13C to 13F show the relationship between the number of contracts and maximum power consumption predicted by the load prediction model 114. 13C is a graph showing the results of model a (conventional method + simple regression analysis). The prediction results are shown in Fig. 13D by model b (simple regression analysis), and Fig. 13E by model c. Figure 13F shows the prediction results by model d (multiple regression analysis), and Figure 13F shows the prediction results by model d (multiple regression analysis). The measurement results are shown respectively.

[0104] Coefficient of determination R 2 A table showing the comparison is shown in Figure 13G. Coefficient of determination R 2 From the values ​​of , model b (simple regression analysis) and model d (multiple regression analysis) It is clear that high accuracy is achieved.

[0105] <Verification of Model b and Model d> For model b (simple regression analysis) and model d (multiple regression analysis), we used a broader period The load prediction model 114 is trained using learning data based on the measured values, and each prediction result is recalculated. It was verified several times.

[0106] Figure 13H shows the prediction results of model b (simple regression analysis) trained using the above training data. FIG. 13I is a graph showing the coefficients and intercepts of each explanatory variable of the above model b. In this example, the coefficient of determination R 2 The result was 0.715, confirming that high accuracy was achieved.

[0107] Figure 13J shows the prediction results of model d (multiple regression analysis) trained using the above training data. FIG. 13K shows the coefficients and intercepts of each explanatory variable in model d. In this example, the coefficient of determination R 2 is 0.810, which is even more accurate than model b. As shown in the figure, in this example, the three explanatory variables ( Number of contracts for supply method (single-phase three-wire 100V), total contracted power for metered electricity B, contracts with water heaters The addition of these explanatory variables also improved accuracy. In addition, it is assumed that the small number of theories that have a high impact on the objective variable By simply adding a clear variable, the processing load (calculation load) can be reduced and accuracy can be improved efficiently. You will know what you can do.

[0108] <Verification of margin calculation method> Next, in order to set the margin appropriately, the number of contracts and the difference ( The relationship between the difference between the predicted value and the actual measured value was verified.

[0109] Figure 14A shows the effect of the number of main switches (or number of subscribers) on model b (simple regression analysis). This is a graph plotting the difference between the above for each period. The figure also shows the interval in which 95% of the predicted values ​​fall. The upper limit (average value + standard deviation σ × 2) is shown in a plot with a different pattern from the other plots. The percentage of predicted values ​​greater than the actual measured values ​​was 57%.

[0110] Figure 14B shows the effect of model d (multiple regression analysis) on the number of main switches (or the number of customers) This is a graph plotting the difference between the above for each period. The figure also shows the interval in which 95% of the predicted values ​​fall. The upper limit (average value + standard deviation σ × 2) is plotted with different patterns. The percentage of values ​​greater than the actual measured value was 59%.

[0111] Figure 14C shows the effect of the number of main switches (or number of subscribers) on model b (simple regression analysis). This is a graph plotting the difference between the above for each group. The upper limit (average value + standard deviation σ × 3) between the two plots is shown by a plot with a different pattern from the other plots. The percentage of predicted values ​​greater than the actual measured values ​​was 57%.

[0112] Figure 14D shows the effect of model d (multiple regression analysis) on the number of main switches (or the number of customers) This is a graph plotting the difference between the above for each group. The upper limit (average value + standard deviation σ × 3) between the two plots is shown by a plot with a different pattern from the other plots. The percentage of predicted values ​​greater than the actual measured values ​​was 59%.

[0113] Next, based on the graphs shown in FIG. 14A to FIG. 14D, the upper limit of the range in which 95% of the predicted value falls is calculated. The limit values ​​were approximated using four types of approximation formulas (linear approximation, exponential approximation, power approximation, and logarithmic approximation).

[0114] FIG. 15A shows an approximation of the upper limit of the interval in which 95% of the predicted values ​​fall based on FIG. 14A. 15B is a graph showing the results. FIG. 15B shows the parameters of each approximate expression for the above results. The percentage of times the predicted values ​​exceed the actual values, and the average of the differences are shown in table format.

[0115] As shown in Figure 15B, the rate at which the predicted value exceeds the actual measured value is 95% for the conventional method. The average of the intervals for each approximation formula was 94%, so the two were roughly the same. Regarding the power consumption, if the approximate formula is not used (hereinafter referred to as the "conventional method"), the power consumption is 5.58 kW. In contrast, when the approximate formula is used, the output is about 3.7kW. It can be seen that the accuracy improves by using

[0116] FIG. 15C shows an approximation of the upper limit of the interval in which 95% of the predicted values ​​fall based on FIG. 14B. FIG. 15D is a graph showing the results. The percentage of measurements that exceed the actual measurements and the average of the differences are shown in table format.

[0117] As shown in Figure 15D, the rate at which the predicted value exceeds the actual measured value is 95% for the conventional method. The average of the intervals for each approximation formula was 94%, so the two were roughly the same. Regarding the conventional method, the figure is 5.58kW, while when the approximate formula is used, the figure is about The result is 3.0 kW, and it can be seen that the accuracy is improved by using the approximation formula.

[0118] FIG. 15E shows an approximation of the upper limit of the interval in which 99.7% of the predicted values ​​fall based on FIG. 14C. FIG. 15F is a graph showing the results of the above calculation. , the percentage of times the predicted value exceeds the actual value, and the average of the above differences are shown in table format.

[0119] As shown in Figure 15F, the rate at which the predicted value exceeds the actual measured value is 95% for the conventional method. In contrast, the average for each approximation method was 98%, meaning that the approximation method outperformed the conventional method. On the other hand, the average difference was 5.58kW for the conventional method, while In both cases, the results were approximately 6.0 kW, and the accuracy was lower when the approximation formula was used.

[0120] Figure 15G shows an approximation of the upper limit of the interval in which 99.7% of the predicted values ​​fall based on Figure 14D. FIG. 15H shows the results of the above calculation. , the percentage of times the predicted value exceeds the actual value, and the average of the above differences are shown in table format.

[0121] As shown in Figure 15H, the average difference is similar between the conventional method and each approximation formula. The percentage of predicted values ​​exceeding the actual values ​​was 95% for the conventional method, whereas the average of the intervals for each approximation method was 9 8%, which shows that the approximation formula can cover more cases. It can be seen that.

[0122] From the above, an approximation formula was generated that shows the relationship between the number of main switches and the difference between the actual measured value and the predicted value. By forming the difference, except for the case of FIG. 15F, the above difference can basically be used as a guide for the tolerance. It was confirmed that it is possible to obtain the value with high accuracy.

[0123] Although the embodiment of the present invention has been described in detail above, the above description is intended to facilitate understanding of the present invention. The present invention is not intended to limit the scope of the present invention. It is of course possible to modify or improve the present invention, and the present invention includes equivalents thereof. For example, the above embodiment has been described in detail in order to clearly explain the present invention. The present invention is not limited to the above-described embodiment. It is possible to add, delete, or replace some of the configurations of the form with other configurations.

[0124] For example, in the above, the load prediction model 114 is described as a model shown in FIG. As explained above, the load prediction model 114 is, for example, a deep neural network (DNN), This can be achieved by other types of models such as Support Vector Machines (SVMs). You may do so. [Explanation of symbols]

[0125] 1 Transformer load forecasting system, 2 Consumer, 3 System center, 4 Power information management device , 5 Communication network, 7 Transformer, 8 Power distribution system, SM Smart meter, MS Main Switch, L load, 100 load forecasting device, 110 memory unit, 111 consumer information, 11 2 Measurement information, 113 Learning data, 114 Load forecasting model, 115 Target transformer for forecasting Transformer information, 116 Load forecast results, 117 Margin calculation results, 118 Transformer installation information, 12 0 consumer information management unit, 130 measurement value acquisition unit, 135 learning data generation unit, 140 model Data learning unit, 145 prediction target transformer information acquisition unit, 150 load prediction unit, 155 margin calculation unit output unit, 160 result presentation unit, S900 model learning process, 1000 learning data setting screen , S1100 Load forecast processing, 1200 Result display screen

Claims

1. The information processing device includes a processor and a storage device. The number of main switches by capacity installed at each customer receiving power from the transformer is A load forecasting model is a model in which the maximum power of the transformer is the objective variable and the explanatory variable is the Generate, Each transformer that receives power from the transformer that is the target of the maximum power prediction is The number of main switches installed at consumers by capacity was used as an explanatory variable in the load forecast model. and determining the objective variable by inputting the objective variable, and outputting the determined objective variable. Transformer load forecasting system.

2. 2. The transformer load prediction system according to claim 1, The power supply is supplied from the transformer to the customer. If there is a main switch whose capacity is unknown, the main switch whose capacity is unknown shall be Determine the capacity based on the contents of the power supply contract of the customer where the device is installed, Transformer load forecasting system.

3. 3. A transformer load prediction system according to claim 1, The load prediction model is a simple regression analysis, a multiple regression analysis, a deep neural network (DNN), and Support Vector Machine (SVM) Generate Transformer load forecasting system.

4. 4. A transformer load prediction system according to claim 3, comprising: In the multiple regression analysis, the number of contracts for the supply system (single-phase three-wire 100V) was used as an explanatory variable. , the total contracted power of metered electricity B, and the number of contracts with water heaters. Use Transformer load forecasting system.

5. 3. A transformer load prediction system according to claim 1, The relationship between the number of main switches receiving power from a transformer and the actual measured value of the maximum power of the transformer is Remember, Based on the load prediction model, the relationship between the number of main switches and the predicted maximum power of the transformer is calculated. Seek, An approximation formula showing a relationship between the number of main switches and a difference between the predicted value and the actual measured value is generated. 、 outputting information indicating the difference calculated using the generated approximation formula; Transformer load forecasting system.

6. An information processing device having a processor and a storage device, The number of main switches by capacity installed at each customer receiving power from the transformer is A load forecasting model is a model in which the maximum power of the transformer is the objective variable and the explanatory variable is the generating Each transformer that receives power from the transformer that is the target of the maximum power prediction is The number of main switches installed at consumers by capacity was used as an explanatory variable in the load forecast model. determining the objective variable by inputting the objective variable, and outputting the determined objective variable; A transformer load forecasting method.

7. 7. A transformer load prediction method according to claim 6, comprising the steps of: The power supply is supplied from the transformer to the customer. If there is a main switch whose capacity is unknown, the main switch whose capacity is unknown shall be determining a capacity based on the contents of a power supply contract of the customer in which the device is installed; The transformer load forecasting method further comprises:

8. 7. A transformer load prediction method according to claim 6, comprising the steps of: The information processing device, The relationship between the number of main switches receiving power from a transformer and the actual measured value of the maximum power of the transformer is A storing step; Based on the load prediction model, the relationship between the number of main switches and the predicted maximum power of the transformer is calculated. The steps required, An approximation formula showing a relationship between the number of main switches and a difference between the predicted value and the actual measured value is generated. and outputting information indicating the difference calculated using the generated approximation formula; A transformer load forecasting method.

Images