Output voltage regulator, output voltage regulator system, output voltage regulator method and program

Abstract

To provide a technique capable of optimally adjusting a transmission voltage of a distribution line.SOLUTION: A transmission voltage adjustment device 200 comprises, as its function: a voltage acquisition part 232 that acquires a value of a transmission voltage in a time sequence; a prediction part 233 that predicts a power consumption amount at a predetermined first time sequence and a value of the transmission voltage at a second time and thereafter that is the predetermined first time and thereafter from the change in the time sequence of the value of the transmission voltage up to the predetermined first time on the basis of a machine learning model by the power consumption amount in the past distribution line; a determination part 235 that determines the value of the transmission voltage on the basis of a plan value of the transmission voltage that is previously stored and the value of the predicted transmission voltage; and a voltage adjustment part 236 that adjusts the transmission voltage of the distribution line to the value of the determined transmission voltage.SELECTED DRAWING: Figure 4

Description

Technical Field

[0007] ,

[0001] The present invention relates to a transmission voltage adjustment device, a transmission voltage adjustment system, a transmission voltage adjustment method, and a program for adjusting the transmission voltage of a distribution line.

Background Art

[0002] The power transmitted by the distribution lines of a substation is adjusted by appropriately adjusting the voltage according to the time of day and season. Conventionally, the voltage adjustment of the distribution lines has been performed by an operator traveling to each substation and manually operating the voltage adjustment device.

[0003] Therefore, a setting value determination device capable of autonomously determining the setting value to be set in the equipment for voltage adjustment of the power system is known (see, for example, Patent Document 1). This setting value determination device predicts the load in consideration of weather information, season, etc.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] By the way, the power consumption of the power system varies every day and every moment according to the situation at that time. Therefore, it is necessary to optimally adjust the transmission voltage of the distribution line.

[0006] Therefore, an object of the present invention is to provide a transmission voltage adjustment device, a transmission voltage adjustment system, a transmission voltage adjustment method, and a program capable of optimally adjusting the transmission voltage of a distribution line.

Means for Solving the Problems

[0007] To solve the above problems, the invention of claim 1 is a transmission voltage adjustment device installed in a power distribution substation and for adjusting the transmission voltage of a power distribution line, comprising: a voltage acquisition unit that acquires the value of the transmission voltage in a time series; a prediction unit that predicts the value of the transmission voltage from a second time onward, which is after the predetermined first time, based on the amount of power consumed at a predetermined first time and the change in the time series of the value of the transmission voltage up to the predetermined first time, based on a machine learning model based on the amount of power consumed in the power distribution line in the past; a determination unit that determines the value of the transmission voltage based on a pre-stored planned value of the transmission voltage and the predicted value of the transmission voltage; and a voltage adjustment unit that adjusts the transmission voltage of the power distribution line to the determined value of the transmission voltage.

[0008] The invention of claim 2 is characterized in that, in the transmission voltage adjustment device described in claim 1, the value of the transmission voltage from the second time onward is predicted from the change in the time series of the value of the transmission voltage up to the time approximately immediately before the adjustment of the transmission voltage of the power distribution line, which is set as the predetermined first time.

[0009] The invention of claim 3 is characterized in that, in the transmission voltage adjustment device described in claim 1, it comprises a learning unit that performs machine learning based on a value including the terminal voltage of the power distribution line after the second time and a value of the actual power consumption after the second time, and updates the machine learning model based on the results of the machine learning.

[0010] The invention of claim 4 is characterized in that, in the transmission voltage adjustment device described in claims 1 to 3, the machine learning model is a supervised machine learning model that uses the amount of power consumed by the distribution line over a predetermined period in the past, time information over the predetermined period, and weather information over the predetermined period as training data.

[0011] The invention of claim 5 is characterized in that, in the transmission voltage adjustment device described in claim 1, the voltage adjustment unit adjusts the transmission voltage of the power distribution line by an operation signal to raise / lower a tap to a voltage regulator.

[0012] The invention of claim 6 is a transmission voltage adjustment system comprising: a transmission voltage regulator connected to a distribution line of a substation for adjusting the transmission voltage of the distribution line; and a calculation device for calculating the adjustment value of the transmission voltage by the transmission voltage regulator, wherein the calculation device comprises: a voltage acquisition unit for acquiring the value of the transmission voltage in a time series; a prediction unit for predicting the value of the transmission voltage from a second time onward, which is after the predetermined first time, based on a machine learning model of past power consumption amounts in the distribution line, the power consumption amount at a predetermined first time and the change in the time series of the value of the transmission voltage up to the predetermined first time; a determination unit for determining the value of the transmission voltage based on a pre-stored planned value of the transmission voltage and the predicted value of the transmission voltage; and a voltage adjustment unit for adjusting the transmission voltage of the distribution line to the determined value of the transmission voltage.

[0013] The invention of claim 7 is a method for adjusting the output voltage, which is executed on a computer having a processor and memory, and is installed in a power distribution substation to adjust the output voltage of a power distribution line, wherein the method is characterized in that the processor performs the steps of: acquiring the output voltage value in a time series; predicting the output voltage value from a second time onward, which is after the predetermined first time, based on the amount of power consumed at a predetermined first time and the change in the time series of the output voltage value up to the predetermined first time, using a machine learning model based on the amount of power consumed at the power distribution line in the past; determining the output voltage value based on a pre-stored planned value of the output voltage and the predicted output voltage value; and adjusting the output voltage of the power distribution line to the determined output voltage value.

[0014] The invention of claim 8 is a program to be executed by a computer having a processor and memory, and to be installed in a power distribution substation to adjust the output voltage of a power distribution line, wherein the program is characterized by causing the processor to perform the following steps: acquiring the output voltage value in a time series; predicting the output voltage value from a second time onward, which is after the predetermined first time, based on a machine learning model of past power consumption amounts in the power distribution line, using the power consumption amount at a predetermined first time and the change in the time series of the output voltage value up to the predetermined first time; determining the output voltage value based on a pre-stored planned value of the output voltage and the predicted output voltage value; and adjusting the output voltage of the power distribution line to the determined output voltage value. [Effects of the Invention]

[0015] According to the invention described in claims 1, 6 to 8, a machine learning model based on the power consumption of a power distribution line is used to predict the value of the transmission voltage from a predetermined second time onward based on the time-series change in the value of the transmission voltage of the power distribution line up to a predetermined first time. The value of the transmission voltage is determined based on a pre-stored planned value of the transmission voltage and the predicted value of the transmission voltage, and the voltage is adjusted to the determined value.

[0016] In other words, based on the changes in the transmission voltage of the distribution line up to a certain point in time, a machine learning model is used to predict the value of the transmission voltage. The transmission voltage from that point onward is then determined and adjusted based on this predicted value and the planned transmission voltage. This makes it possible to optimally adjust the transmission voltage of the distribution line. As a result, optimal and stable power supply in the power system becomes possible.

[0017] According to the invention described in claim 2, the value of the transmission voltage is predicted from the time series change of the transmission voltage value at approximately the time immediately before the adjustment of the transmission voltage of the distribution line, i.e., in near real-time. This makes it possible to optimally adjust the transmission voltage of the distribution line based on the load of the power system in real time.

[0018] According to the invention described in claim 3, machine learning is performed based on a value including the terminal voltage of the distribution line after the second time and a value of the actual power consumption amount after the second time, and the machine learning model is updated based on the result of the machine learning. That is, further machine learning is performed based on a value including the terminal voltage of the distribution line after adjusting the predicted transmission voltage value, that is, a value at a representative point such as the terminal voltage, and a value of the actual power consumption amount, and the machine learning model is updated. Thereby, it becomes possible to predict a refined transmission voltage value based on the load of the latest power system.

[0019] According to the invention described in claim 4, the machine learning model used for predicting the value of the transmission voltage is a machine learning model that uses the power consumption amount by the distribution line in a past predetermined period, time information in the predetermined period, and weather information in the predetermined period as teacher data. That is, not only the power consumption amount but also time information and weather information are used as the teacher data of the machine learning model. Thereby, it becomes possible to predict a refined transmission voltage value in consideration of various elements in the latest power supply.

[0020] According to the invention described in claim 5, the transmission voltage of the distribution line is adjusted by an operation signal for raising / lowering the tap for the voltage regulator. Therefore, it becomes possible to apply the present invention without replacing the existing voltage regulator. Thereby, it becomes possible to configure the adjustment of the transmission voltage without incurring more cost.

Brief Description of the Drawings

[0021] [Figure 1] It is a circuit configuration diagram showing the circuit configuration of the voltage regulator 100 according to the conventional embodiment. [Figure 2] It is a graph showing an example of a daily load curve that is the basis of voltage adjustment. [Figure 3] It is a circuit configuration diagram showing the circuit configuration of the transmission voltage adjustment device 200 according to Embodiment 1 of the present invention. [Figure 4] It is a functional block diagram showing the functions of the transmission voltage adjustment device 200 in FIG. 3. [Figure 5]It is a flowchart showing the control procedure by the control unit 230 in FIG. 4. [Figure 6] It is a block diagram showing the procedures of machine learning and prediction by the control unit 230 in FIG. 4. [Figure 7] It is a functional block diagram showing the computer 700 according to Embodiment 2 of the present invention.

Embodiments for Carrying Out the Invention

[0022] Hereinafter, this invention will be described based on the illustrated embodiments.

[0023] (Overview) The transmission voltage adjustment device according to the embodiment of the present invention is a device for adjusting the voltage transmitted from a substation to a distribution line to an appropriate voltage according to the power consumption that varies depending on the time zone and season. The transmission voltage adjustment device adjusts the transmission voltage to an appropriate value by outputting an operation signal to a voltage regulator for adjusting the voltage from the 6kV bus of a distribution substation.

[0024] FIG. 1 is a circuit configuration diagram showing the circuit configuration of a voltage regulator 100 according to a conventional embodiment. The voltage regulator 100 shown in FIG. 1 is connected to, for example, a high-voltage distribution line that is a three-phase alternating current consisting of phases A, B, and C, and is a device for adjusting the voltage of the high-voltage distribution line. An operation mechanism 101 is provided therefor. The operation mechanism 101 is configured to be operated by an operation signal 400 from a voltage relay 300. The operation mechanism 101 is configured by, for example, a load tap changer, and is configured to adjust by raising / lowering a tap (TaP) for adjusting the bus voltage. The voltage regulator 100 and the voltage relay 300 shown in FIG. 1 constitute a transmission voltage adjustment system 10.

[0025] The voltage relay 300 is a device that measures the voltage from the 6kV busbar of the distribution substation and outputs an operation signal 400 to the operating mechanism 101 in order to adjust the output voltage. The voltage relay 300 is configured to measure, for example, the line voltage between phases A and C on the load side, and measures the line voltage between phases A and C using electrical signals output from the primary high-voltage distribution lines VA and VC of the voltage regulator 100 to the secondary voltage signal line. The voltage relay 300 also outputs an operation signal 400 to adjust the output voltage, either manually or by program control, based on the measurement results.

[0026] Figure 2 is a graph showing an example of a daily load curve that forms the basis for voltage adjustment. The graph in Figure 2 shows the time change of the set value of the output voltage over a day, with the vertical axis representing the output voltage (in V) and the horizontal axis representing the time over a day, and the set value of the output voltage is shown for each time period as shown by curve L1. Voltage adjustment by the voltage relay 300 is performed based on the value of curve L1, which shows the change in the value of the daily load curve shown in Figure 2. For example, during periods other than summer and winter, the set value of the output voltage is set to 6,600V throughout the day, but during summer and winter, as shown by curve L1, it is set to 6,600V from 0:00 to 6:00 and from 23:00 to 0:00, set to 6,630V from 6:00 to 17:00, and set to 6,660V from 17:00 to 23:00. In this way, the set value of the output voltage is set to change according to the season and time of day.

[0027] However, as mentioned above, the amount of electricity used in the power grid fluctuates constantly depending on the circumstances at the time, specifically weather conditions such as temperature and sunshine hours, as well as the time of day and whether it is a weekday or a holiday. Therefore, it is necessary to optimally adjust the transmission voltage of the distribution lines. It is believed that such optimization of the transmission voltage of distribution lines can be accurately predicted using machine learning based on information such as past electricity usage and weather conditions.

[0028] Therefore, in this embodiment, based on a machine learning model that uses the power consumption of the distribution line, the value of the transmission voltage is predicted from the time-series change in the value of the transmission voltage of the distribution line up to a certain time. The value of the transmission voltage is determined based on the pre-stored planned value of the transmission voltage of the daily load curve and the predicted value of the transmission voltage, and the voltage is adjusted to the determined voltage. This makes it possible to optimally adjust the transmission voltage of the distribution line.

[0029] (Embodiment 1) <Structure> Figures 3 to 6 illustrate this embodiment, and Figure 3 is a circuit diagram showing the circuit configuration of the transmission voltage adjustment device 200 according to Embodiment 1 of the present invention. The voltage regulator 100 shown in Figure 3 is connected to a high-voltage distribution line which is a three-phase AC consisting of A phase, B phase, and C phase, and is a device for adjusting the voltage of the high-voltage distribution line. The transmission voltage adjustment device 200 shown in Figure 3 is a device installed in place of the voltage relay 300 shown in Figure 1, and is similar to the voltage relay 300 in that it measures the voltage from the 6kV busbar of the distribution substation and outputs an operation signal 400 to the operation mechanism 101 in order to adjust the transmission voltage. The transmission voltage adjustment device 200 is configured to measure the line voltage between each of the A phase, B phase, and C phase on the load side, and measures the line voltage between each of the A phase, B phase, and C phase by an electrical signal output from the secondary voltage signal line from the primary high-voltage distribution line of the voltage regulator 100. Furthermore, when measuring the line voltage, the output voltage regulator 200 may directly measure the line voltage input via a wire, or it may read and acquire the reading of a voltmeter installed on a metering panel or the like from image data of the voltmeter, or it may acquire the reading as a signal indicating the reading via a transducer.

[0030] Furthermore, the difference between the output voltage regulator 200 and the voltage relay 300 is that it is configured to measure the currents of phases A, B, and C on the load side, and as shown in Figure 3, for example, it measures currents Ia, Ib, and Ic. When measuring current, the output voltage regulator 200 may measure the current input directly via a wire, or it may acquire the indicated value of an ammeter installed on a metering panel, etc., by reading it from image data of the ammeter, etc., or it may acquire it as a signal indicating the indicated value via a transducer. The voltage regulator 100 and the output voltage regulator 200 shown in Figure 3 constitute the output voltage regulation system 1.

[0031] The output voltage regulator 200 is composed of, for example, a computer (desktop, laptop, tablet, etc.) or a server device. The output voltage regulator 200 may be configured to operate by being located in a substation and communicating with terminal devices via a network, or it may be located in a location separate from the substation and configured to output an operation signal 400 to the operation mechanism 101 via remote control.

[0032] Figure 4 is a functional block diagram showing the functions of the transmission voltage regulator 200 shown in Figure 3. The transmission voltage regulator 200 comprises a communication unit 210, a storage unit 220, and a control unit 230. As shown in Figure 4, the communication unit 210, the storage unit 220, and the control unit 230 are electrically connected by a bus or the like.

[0033] The communication unit 210 is a communication interface for wired or wireless communication in order to output an operation signal 400 to the operating mechanism 101, and any communication protocol can be used as long as communication between them is possible. This communication unit 210 communicates using a communication protocol such as TCP / IP (Transmission Control Protocol / Internet Protocol).

[0034] The memory unit 220 stores programs and input data for executing various control processes and functions within the control unit 230, and consists of memory including RAM (Random Access Memory) and ROM (Read Only Memory), and storage including HDD (Hard Disk Drive), SSD (Solid State Drive), and DRAM (Dynamic Random Access Memory). The memory unit 220 also stores the daily load curve database 221 and the machine learning model 222. Furthermore, The daily load curve database 221 stores daily load curve information (planned values ​​of the transmission voltage) as shown in Figure 2. Daily load curve information refers to information showing the set values ​​of the transmission voltage for each hour of the day, and for example, each value in the graph shown in Figure 2 is stored. The daily load curve information stored in the daily load curve database 221 may include past daily load curve information and future (planned) daily load curve information, and may be stored by date, by date when the values ​​fluctuate, by season, etc.

[0035] Machine learning model 222 is a machine learning model for predicting the value of the transmission voltage, and is a prediction algorithm that has been trained (learned) to recognize a specific type of pattern in the value of the transmission voltage. For example, machine learning model 222 is a machine learning model based on the amount of power consumption in a power distribution line, and is a machine learning model that has learned to recognize the pattern between past power consumption information and the value of the transmission voltage at that time. Furthermore, machine learning model 222 may be a machine learning model that has been trained based on values ​​at representative points such as the terminal voltage of a power distribution line, in addition to past power consumption, or it may be a machine learning model that has been trained based on power consumption over a predetermined period in the past, information about that period (period information), and weather information during that period (used as training data). Machine learning model 222 may be generated by machine learning performed as a function of the learning unit 234 described later, or it may be a machine learning model that has been generated as a result of machine learning performed by another device. The machine learning for generating the machine learning model will be described later.

[0036] The control unit 230 controls the overall operation of the output voltage regulator 200 by executing a program stored in the memory unit 220, and is composed of devices including a CPU (Central Processing Unit), MPU (Micro Processing Unit), GPU (Graphics Processing Unit), microprocessor, processor core, multiprocessor, ASIC (Application-Specific Integrated Circuit), and FPGA (Field Programmable Gate Array). The functions of the control unit 230 include a daily load curve information acquisition unit 231, a voltage acquisition unit 232, a prediction unit 233, a learning unit 234, a determination unit 235, and a voltage adjustment unit 236. The daily load curve information acquisition unit 231, the voltage acquisition unit 232, the prediction unit 233, the learning unit 234, the determination unit 235, and the voltage adjustment unit 236 are activated by a program stored in the memory unit 220 and executed by the output voltage regulator 200.

[0037] The daily load curve information acquisition unit 231 acquires information on the daily load curve, which is the planned value of the transmission voltage. For example, the daily load curve information is input by the person in charge of adjusting the transmission voltage through an input operation from the transmission voltage adjustment device 200 or terminal device, and the daily load curve information acquisition unit 231 accepts the input information. The daily load curve information is information that shows the set value of the transmission voltage for each hour of the day, as described above. In addition, if, for example, daily load curve information for the relevant date does not exist, the daily load curve information acquisition unit 231 may output warning information to notify the person in charge of adjusting the transmission voltage.

[0038] The voltage acquisition unit 232 acquires the values ​​of the transmission voltage of the distribution line in a time series. Specifically, as an example, the voltage acquisition unit 232 acquires the values ​​of the line voltages between phases A, B, and C, measured from the electrical signals output from the secondary voltage signal line shown in Figure 3.

[0039] The prediction unit 233 predicts the value of the transmission voltage from a second time onward, which is after a predetermined first time, based on a machine learning model 222 that uses past power consumption data from the power distribution line, using the power consumption up to a predetermined first time and the time-series change in the value of the transmission voltage. The predetermined first time is, for example, approximately immediately before the transmission voltage is adjusted, and the value of the transmission voltage up to the predetermined first time refers to the transmission voltage in approximate real time. That is, the prediction unit 233 predicts the value of the transmission voltage from the time-series change in the value of the transmission voltage over a predetermined time range, starting from the approximately real-time value of the transmission voltage. However, the predetermined first time is not limited to the above. The value of the transmission voltage at this time is the value acquired by the voltage acquisition unit 232. The second time onward, which is after the predetermined first time, is, for example, the timing of the transmission voltage adjustment, and the prediction unit 233 predicts the value of the transmission voltage from this timing onward. However, similar to the predetermined first time, the predetermined second time is not limited to the above.

[0040] The prediction of the output voltage value by the prediction unit 233 based on the machine learning model 222 is performed as follows: The prediction unit 233 recognizes the pattern of the output voltage value stored in the machine learning model 222 from the amount of power consumed at the time immediately before adjusting the output voltage and the time-series change in the output voltage value up to the time immediately before adjusting the output voltage. Then, the prediction unit 233 predicts the output voltage value after the time of adjustment from the information showing the change in the output voltage value in that pattern and calculates the predicted output voltage value.

[0041] The power consumption in the prediction unit 233 is calculated, for example, by the product of the line voltages Vab, Vbc, and Vca between the A, B, and C phases, which are output from the secondary voltage signal line, and the currents Ia, Ib, and Ic of the A, B, and C phases, respectively.

[0042] The learning unit 234 performs machine learning based on the values ​​of representative points such as the terminal voltage of the distribution line (values ​​including the terminal voltage) from the second time point onward, and the actual power consumption values ​​from the second time point onward, and updates the machine learning model 222 based on the results of the machine learning. The values ​​of representative points such as the terminal voltage of the distribution line are, for example, the phase voltages Va, Vb, and Vc between the A, B, and C phases output from the secondary voltage signal line, and are values ​​acquired by the voltage acquisition unit 232. The actual power consumption values ​​are calculated, for example, by the same calculation as the power consumption calculation in the prediction unit 233.

[0043] The machine learning in the learning unit 234 may be performed by supervised machine learning using predetermined training data, by unsupervised machine learning, or by deep learning. In this embodiment 1, it will be described as supervised machine learning using the values ​​of representative points such as the terminal voltage of a power distribution line and the actual power consumption values ​​as training data. The learning unit 234 may also perform data processing (annotation) on the training data, such as the values ​​of representative points such as terminal voltage and the power consumption values, by adding, for example, the interpretation indicated by each value as tag information. Furthermore, when updating the machine learning model 222, the learning unit 234 may update it by retraining, or it may be configured to update only when the performance of the machine learning model 222 has improved after a performance evaluation.

[0044] The determination unit 235 determines the output voltage value based on the pre-stored planned output voltage value and the predicted output voltage value. The pre-stored planned output voltage value is specifically the daily load curve information stored in the daily load curve database 221, which is the set value of the output voltage for each hour of the day. The predicted output voltage value is the predicted output voltage value calculated by the prediction unit 233.

[0045] The determination of the output voltage value by the determination unit 235 is performed as follows: The determination unit 235 calculates the difference between the predicted output voltage value calculated by the prediction unit 233 and the set value of the output voltage based on the daily load curve. Then, it determines how to raise / lower the tap (TaP) of the load-dependent tap switching device in the operating mechanism 101.

[0046] The voltage adjustment unit 236 adjusts the output voltage of the distribution line to a determined output voltage value. The determined output voltage is the output voltage value determined by the determination unit 235. Specifically, the voltage adjustment unit 236 adjusts the output voltage of the distribution line by transmitting an operation signal 400 to the operation mechanism 101, which controls the raising / lowering of the tap (TaP) of the load tap switching device determined by the determination unit 235.

[0047] <Processing flow> Referring to Figure 5, an example of the control process (transmission voltage adjustment method) by the control unit 230 of the transmission voltage adjustment device 200 will be explained. Figure 5 is a flowchart showing the control procedure by the control unit 230 in Figure 4.

[0048] As part of the process in step S101, the voltage acquisition unit 232 of the control unit 230 acquires the values ​​of the output voltage of the power distribution line up to the present in a time series. In step S101, for example, the measured values ​​of the phase voltages Va, Vb, and Vc of phases A, B, and C, output from the secondary voltage signal line, are acquired in a time series.

[0049] As part of the processing in step S102, the prediction unit 233 of the control unit 230 predicts the value of the transmission voltage after the timing of transmission voltage adjustment, based on the time-series change in the transmission voltage value up to approximately immediately before the transmission voltage adjustment, which was obtained in step S101, using the machine learning model 222. In step S102, for example, the pattern of the transmission voltage value stored in the machine learning model 222 is recognized from the power consumption at the timing immediately before the transmission voltage adjustment and the time-series change in the transmission voltage value up to the timing immediately before the transmission voltage adjustment. Then, the transmission voltage value after the timing of transmission voltage adjustment is predicted from the information showing the change in the transmission voltage value in that pattern, and the predicted value of the transmission voltage is calculated.

[0050] As part of step S103, the determination unit 235 of the control unit 230 reads the daily load curve information stored in the daily load curve database 221 and obtains the set value of the transmission voltage for each hour of the day.

[0051] In step S104, the determination unit 235 of the control unit 230 determines the value of the transmission voltage based on the hourly transmission voltage setting value for the day (pre-stored planned transmission voltage value) obtained in step S103 and the transmission voltage value predicted in step S102.

[0052] In step S105, the voltage adjustment unit 236 of the control unit 230 adjusts the acquired output voltage of the power distribution line to the output voltage value determined in step S104. In step S105, for example, the output voltage of the power distribution line is adjusted by sending an operation signal 400 to the operation mechanism 101 to raise or lower the tap (TaP) of the load tap switching device determined by the determination unit 235.

[0053] Referring to Figure 6, the machine learning and prediction procedure performed by the control unit 230 of the output voltage regulator 200 will be explained. Figure 6 is a flowchart showing the machine learning and prediction procedure performed by the control unit 230 in Figure 4.

[0054] As part of the processing in step S201, the learning unit 234 of the control unit 230 collects training data for machine learning. Examples of training data include past power consumption amounts in the distribution line, information indicating external factors related to said power consumption, and the value of the transmission voltage of the distribution line. Past power consumption amounts in the distribution line are calculated as described above by the product of the line voltages Vab, Vbc, Vca between the A, B, and C phases output from the secondary voltage signal line and the currents Ia, Ib, Ic of the A, B, and C phases, respectively. Information indicating external factors related to said power consumption includes information on the timing of said power consumption, specifically the date, day of the week, time of day, etc., and also includes seasonal information obtained from the date. Furthermore, information indicating external factors related to said power consumption includes weather information at the time of said power consumption, specifically the weather (clear, sunny, cloudy, rainy, etc.) and temperature information. As described above, the output voltage values ​​of the distribution line are the line voltages Vab, Vbc, and Vca between phases A, B, and C, respectively, which are output from the secondary voltage signal line.

[0055] As part of the process in step S202, the learning unit 234 of the control unit 230 performs supervised machine learning using the training data collected in step S201 to generate or update the machine learning model 222.

[0056] As part of step S203, the prediction unit 233 of the control unit 230 collects input data for predicting the value of the transmission voltage based on the machine learning model 222. Examples of input data include the value of the transmission voltage up to approximately immediately before (currently) adjusting the transmission voltage, and the value of power consumption.

[0057] As part of the processing in step S204, the prediction unit 233 of the control unit 230 predicts the value of the transmission voltage after the timing for adjusting the transmission voltage, based on the input data collected in step S203.

[0058] As part of the processing in step S205, the prediction unit 233 of the control unit 230 outputs the prediction result from step S204.

[0059] <Effects> According to the first embodiment of the transmission voltage adjustment device 200 and transmission voltage adjustment system 1, based on past power consumption information and a machine learning model 222 that uses the transmission voltage value at that time, the transmission voltage value is predicted from the time-series change in the transmission voltage value of the distribution line up to a certain time. The transmission voltage value is determined based on the planned transmission voltage value of the daily load curve stored in the daily load curve database 221 and the predicted transmission voltage value, and the voltage is adjusted to the determined voltage. As a result, it becomes possible to optimally adjust the transmission voltage of the distribution line. This enables the optimal and stable supply of power in the power system.

[0060] Furthermore, the transmission voltage adjustment device 200 and the transmission voltage adjustment system 1 predict the transmission voltage value from the time-series changes in the transmission voltage value in near real-time. This makes it possible to optimally adjust the transmission voltage of the distribution line based on the real-time load of the power system.

[0061] Furthermore, machine learning is performed based on representative values ​​such as the terminal voltage of the distribution line and the actual power consumption values ​​after the timing of the transmission voltage adjustment, and the machine learning model is updated based on the results of the machine learning. In other words, machine learning is performed again based on representative values ​​such as the terminal voltage of the distribution line after the transmission voltage has been adjusted to the predicted value and the actual power consumption values, and the machine learning model is updated. This makes it possible to predict the transmission voltage value precisely based on the load of the power system in the immediate vicinity.

[0062] (Embodiment 2 (Program)) Figure 7 is a functional block diagram showing an example of the configuration of a computer (electronic computer) 700 according to Embodiment 2. The computer 700 includes a CPU 701, a main memory 702, an auxiliary memory 703, and an interface 704.

[0063] Here, we will describe in detail the programs for realizing each function that constitutes the daily load curve information acquisition unit 231, voltage acquisition unit 232, prediction unit 233, learning unit 234, determination unit 235, and voltage adjustment unit 236 according to Embodiment 1. These functional blocks are implemented in the computer 700. The operation of each of these components is stored in auxiliary storage device 703 in the form of a program. The CPU 701 reads the program from auxiliary storage device 703, loads it into main memory device 702, and executes the above-mentioned processing according to the program. The CPU 701 also allocates memory areas in main memory device 702 corresponding to the above-mentioned memory units according to the program.

[0064] Specifically, the program is a program that uses a computer 700 to implement the following steps: acquiring the value of the transmission voltage in a time series; predicting the value of the transmission voltage from a second time onward (after a predetermined first time) based on the amount of power consumed at a predetermined first time and the time series change of the value of the transmission voltage up to the predetermined first time, using a machine learning model based on the amount of power consumed at a predetermined first time and the value of the transmission voltage up to the predetermined first time; determining the value of the transmission voltage based on a pre-stored planned value of the transmission voltage and the predicted value of the transmission voltage; and adjusting the transmission voltage of the distribution line to the determined value of the transmission voltage.

[0065] The auxiliary storage device 703 is an example of a non-temporary tangible medium. Other examples of non-temporary tangible media include magnetic disks, magneto-optical disks, CD-ROMs, DVD-ROMs, and semiconductor memory connected via the interface 704. Furthermore, if this program is distributed to the computer 700 via a network, the receiving computer 700 may expand the program into the main memory 702 and execute the above-described process.

[0066] Furthermore, the program may be intended to implement some of the functions described above. In addition, the program may be a so-called differential file (differential program) that implements the functions described above in combination with other programs already stored in the auxiliary storage device 703.

[0067] Although embodiments of this invention have been described in detail above, the specific configuration is not limited to these embodiments, and any design changes, etc., that do not depart from the gist of this invention are also included. For example, in the above embodiment, a voltage output regulator 200 is provided instead of a voltage relay 300, but a configuration in which a voltage output regulator 200 is provided in addition to a voltage relay 300, and the voltage output regulator 200 is provided to complement the function of the voltage relay 300, is also possible. [Explanation of Symbols]

[0068] 1.10: Output voltage adjustment system 100: Voltage regulator 101: Operation mechanism 200: Output voltage regulator 210: Communications Department 220: Storage section 221: Daily Load Curve Database 222: Machine Learning Models 230: Control Unit 231: Daily load curve information acquisition section 232: Voltage acquisition unit 233: Prediction Department 234: Learning Department 235: Decision Section 236: Voltage Regulating Unit 300: Voltage relay 400: Operation signal 700: Computer 701: CPU 702: Main memory 703 :Auxiliary storage device 704: Interface Va, Vb, Vc: Phase voltage Vab, Vbc, Vca: Line voltages Ia,Ib,Ic: Line current

Claims

1. A voltage regulator installed in a power distribution substation that adjusts the voltage of the power distribution lines, A voltage acquisition unit that acquires the value of the aforementioned transmission voltage in a time series, A prediction unit predicts the value of the transmission voltage from a second time onward, which is after the predetermined first time, based on a machine learning model using past power consumption data for the aforementioned power distribution line, using the power consumption at a predetermined first time and the time-series change in the value of the transmission voltage up to the predetermined first time. A determination unit that determines the value of the transmission voltage based on a pre-stored planned value of the transmission voltage and a predicted value of the transmission voltage, The system includes a voltage adjustment unit that adjusts the output voltage of the distribution line to a determined output voltage value. A voltage regulator characterized by its features.

2. The prediction unit predicts the value of the transmission voltage from the second time onward based on the time-series change in the value of the transmission voltage up to approximately the time immediately before the adjustment of the transmission voltage of the power distribution line, which is the predetermined first time. The output voltage adjustment device according to feature 1.

3. The system includes a learning unit that performs machine learning based on a value including the terminal voltage of the power distribution line after the second time and the actual power consumption value after the second time, and updates the machine learning model based on the results of the machine learning. The output voltage adjustment device according to feature 1.

4. The aforementioned machine learning model is a machine learning model that uses the amount of electricity consumed by the power distribution line over a predetermined period in the past, time information over the predetermined period, and weather information over the predetermined period as training data. The output voltage adjustment device according to any one of claims 1 to 3.

5. The voltage adjustment unit adjusts the output voltage of the power distribution line based on an operation signal to raise / lower the taps of the voltage regulator. The output voltage adjustment device according to feature 1.

6. A voltage regulator connected to the distribution line of a substation, which adjusts the output voltage of the distribution line, A transmission voltage adjustment system comprising a calculation device for calculating the adjustment value of the transmission voltage by the transmission voltage regulator, The aforementioned computing device is A voltage acquisition unit that acquires the value of the aforementioned transmission voltage in a time series, A prediction unit predicts the value of the transmission voltage from a second time onward, which is after the predetermined first time, based on a machine learning model using past power consumption data for the aforementioned power distribution line, using the power consumption at a predetermined first time and the time-series change in the value of the transmission voltage up to the predetermined first time. A determination unit that determines the value of the transmission voltage based on a pre-stored planned value of the transmission voltage and a predicted value of the transmission voltage, The system includes a voltage adjustment unit that adjusts the output voltage of the distribution line to a determined output voltage value. A transmission voltage adjustment system characterized by the following features.

7. A method for adjusting the output voltage of a power distribution line, which is executed on a computer having a processor and memory and installed in a power distribution substation, The above method involves the processor, The steps include acquiring the value of the aforementioned transmission voltage in a time series, Based on a machine learning model using past power consumption data for the aforementioned power distribution line, the steps include predicting the value of the transmission voltage from a second time onward (after the predetermined first time) based on the power consumption at a predetermined first time and the time-series change in the value of the transmission voltage up to the predetermined first time, The steps include determining the value of the transmission voltage based on a pre-stored planned value of the transmission voltage and a predicted value of the transmission voltage, The steps include adjusting the output voltage of the distribution line to the determined output voltage value, and performing the following: A method for adjusting the output voltage, characterized by the following features.

8. A program to be executed on a computer equipped with a processor and memory, and installed in a power distribution substation to adjust the output voltage of power distribution lines, The program is provided to the processor: The steps include acquiring the value of the aforementioned transmission voltage in a time series, Based on a machine learning model using past power consumption data for the aforementioned power distribution line, the steps include predicting the value of the transmission voltage from a second time onward (after the predetermined first time) based on the power consumption at a predetermined first time and the time-series change in the value of the transmission voltage up to the predetermined first time, The steps include determining the value of the transmission voltage based on a pre-stored planned value of the transmission voltage and a predicted value of the transmission voltage, The steps of adjusting the output voltage of the distribution line to the determined output voltage value are performed. A program characterized by the following features.

Images