Power inference device, power control system, system stabilization system, inference model creation device, and power inference method

Abstract

The purpose of the present disclosure is to use a power storage device for a plurality of purposes. A power inference device (401) according to the present disclosure comprises an acquisition unit (41), a system simulation unit (42), a learning unit (43), and an inference unit (44). The acquisition unit (41) acquires system information. The system simulation unit (42) calculates a main control required output corresponding to the system information by simulating the state of a power system (100) on the basis of the system information. The learning unit (43) creates an inference model capable of inferring the main control required output from arbitrary system information on the basis of a plurality of sets of the corresponding system information and the main control required output. The inference unit (44) inputs the system information at the time of inference to the inference model and infers the main control required output at the time of inference. The output of the power storage device (201) in sub-control is determined on the basis of the remainder obtained by subtracting the inferred main control required output from the maximum output.

Description

Power inference device, power control system, power system stabilization system, inference model creation device, and power inference method

[0001] The present disclosure relates to a technique for estimating the charge / discharge power required for a power storage device connected to a power grid.

[0002] An energy storage device connected to a power grid can increase its utilization rate by charging and discharging to the power grid under control based on multiple different objectives. Here, the multiple controls for the energy storage device have priorities, with the control with the highest priority being called primary control and the control with the lowest priority being called secondary control.

[0003] In this case, the energy storage device must always ensure the charge / discharge power required for the primary control and execute the secondary control with the remaining power. To do this, it is necessary to estimate the charge / discharge power required for the primary control, taking into account the constantly changing state of the power grid.

[0004] Patent Document 1 describes a technology in which the gain of Δf feedback control of a storage battery that performs frequency control is changed so that the remaining capacity of the storage battery does not approach the upper or lower limits, and a system simulation is performed to estimate the future charge / discharge amount or remaining capacity and adjust the gain. A storage battery is an example of a power storage device.

[0005] Japanese Patent Application Laid-Open No. 2017-99131

[0006] The technology described in Patent Document 1 does not assume that a storage battery will be controlled for multiple purposes. The present disclosure aims to use a power storage device for multiple purposes.

[0007] The power inference device disclosed herein infers a required main control output, which is the charging / discharging power required for main control of a power storage device that charges / discharges power to / from a power grid, by combining main control with secondary control that is implemented with a lower priority than the main control based on a control purpose different from that of the main control. The power inference device includes an acquisition unit, a system simulation unit, a creation unit, and an inference unit. The acquisition unit acquires system information representing the state of the power grid. The system simulation unit calculates the required main control output corresponding to the acquired system information by simulating the state of the power grid based on the acquired system information. The creation unit creates an inference model that can infer the required main control output from any system information based on multiple pairs of corresponding system information and required main control output. The inference unit inputs system information at the time of inference into the inference model and infers the required main control output at the time of inference. The output of the power storage device under secondary control is determined based on the remainder obtained by subtracting the estimated required main control output from maximum output.

[0008] According to the power inference device of the present disclosure, a required output for primary control is inferred. The output of the power storage device in secondary control is determined based on the remainder obtained by subtracting the required output for primary control from the maximum output, thereby making it possible to use the power storage device for secondary control while ensuring the output required for primary control. Objects, features, aspects, and advantages of the present disclosure will become more apparent from the following detailed description and the accompanying drawings.

[0009] 1 is a diagram showing the configuration of a power system according to a first embodiment. FIG. 2 is a diagram showing the priority order of primary control and secondary control. FIG. 3 is a block diagram showing the configuration of a power inference device according to a first embodiment. FIG. 4 is a diagram showing an example of the result of a detailed stability calculation assuming that system stabilization is performed by a power storage device. FIG. 5 is a diagram showing an example of the result of a detailed stability calculation assuming that system stabilization is performed by a power storage device. FIG. 6 is a diagram showing an example of the result of a detailed stability calculation assuming that system stabilization is performed by a power storage device. FIG. 7 is a diagram showing an example of the result of a detailed stability calculation assuming that system stabilization is performed by a power storage device. FIG. 8 is a diagram showing the flow of processing in a first input / output pattern of the power inference device according to the first embodiment. FIG. 9 is a diagram showing the flow of processing in a second input / output pattern of the power inference device according to the first embodiment. FIG. 10 is a block diagram showing the configuration of a power control device according to the first embodiment. FIG. 11 is a flowchart showing the pre-learning processing of the power inference device according to the first embodiment. FIG. 12 is a flowchart showing the inference processing of the power inference device according to the first embodiment. FIG. 13 is a diagram showing the output required for stabilization obtained from the result of a detailed stability calculation by the power inference device according to the second embodiment. FIG. 14 is a flowchart showing the pre-learning processing of the power inference device according to the second embodiment. FIG. 15 is a block diagram showing the configuration of a power inference device according to a third embodiment. FIG. 16 is a flowchart showing the pre-learning processing of the power inference device according to the third embodiment. 1 is a diagram illustrating a hardware configuration of a power inference device and a power control device, and FIG. 2 is a diagram illustrating a hardware configuration of a power inference device and a power control device.

[0010] <A. Embodiment 1> <A-1. Overall Configuration> Fig. 1 is a diagram showing a power control system 450 and other configurations according to embodiment 1. The power control system 450 is configured to include a power storage device 201 connected to a power grid 100, a power control device 301, and a power inference device 401.

[0011] The power system 100 is connected to loads 101, 102, 103, and 104, generators 111 and 112, transformers 151 and 152, and inverter power supplies 121 and 122. The generators 111 and 112 are, for example, nuclear power generators, thermal power generators, or hydroelectric power generators. The inverter power supplies 121 and 122 are, for example, generators with inverters, such as solar power generators or wind power generators. Measurement devices 131, 132, 133, 134, 135, and 136 constantly measure the load amounts, power generation amounts, node voltages, branch currents, and the like of the loads 101, 102, 103, and 104, the generators 111 and 112, and the inverter power supplies 121 and 122, and store these measurement values.

[0012] The measuring devices 131, 132, 133, 134, 135, and 136 are connected to a communication network 170. The control devices 141, 142, 143, 144, 145, 146, 147, and 148, which are also connected to the communication network 170, can acquire measured values ​​of the measuring devices 131, 132, 133, 134, 135, and 136 via the communication network 170, and can control the loads 101, 102, 103, and 104, the generators 111 and 112, and the inverter power supplies 121 and 122 based on the measured values. Note that the measured values ​​may be acquired by the control devices 141, 142, 143, 144, 145, 146, 147, and 148 via the communication network 170 without being stored in the measuring devices 131, 132, 133, 134, 135, and 136.

[0013] Furthermore, a power storage device 201 is connected to the power system 100. The power storage device 201 is typically a storage battery, but may also be a flywheel, an electric double layer capacitor, a FACTS device, or the like. The power storage device 201 is controlled by a power control device 301. The power control device 301 constantly measures and stores measurement information related to the power storage device 201. The power control device 301 is also connected to a communication network 170, and is capable of transmitting the measurement information related to the power storage device 201 to control devices 141, 142, 143, 144, 145, 146, 147, 148, and the like via the communication network 170.

[0014] The power inference device 401 is connected to the communication network 170 and can acquire the various measurement values ​​described above. The power inference device 401 can also transmit and receive information to and from the power control device 301 via the communication network 170. Based on the system simulation results and various measurement information, the power inference device 401 calculates the output of the power storage device 201 required to obtain a sufficient control effect of the main control (hereinafter referred to as "main control required output") and transmits this information to the power control device 301.

[0015] Based on the information on the required output for main control obtained from the power inference device 401, the power control device 301 secures the required output for main control of the power storage device 201, and then performs secondary control using the remaining output of the power storage device 201.

[0016] The power storage device 201 charges or discharges (hereinafter referred to as "charge / discharge") between the power storage device 201 and the power grid 100 through a combination of primary control and secondary control. The primary control and secondary control are based on different control purposes, and the secondary control has a lower priority than the primary control. Figure 2 is a diagram showing the concepts of primary control and secondary control.

[0017] In FIG. 2 , the required output for primary control is P1, and the output of the power storage device 201 required to fully obtain the control effect of secondary control (hereinafter referred to as the "required output for secondary control") is P2. The power control device 301 secures the output P1 of the power storage device 201 in advance as the required output for primary control. The power control device 301 also allocates the output P2' of the power storage device 201 that can be used for secondary control (hereinafter referred to as the usable output for secondary control) within the range of the control surplus (PA-P1), which is the difference between the maximum output (rated output PA) of the power storage device 201 and the required output for primary control P1. As shown in FIG. 2 , when the required output for secondary control P2 exceeds the control surplus (PA-P1), the control surplus (PA-P1) becomes the upper limit of the usable output for secondary control P2', and the usable output for secondary control falls short of the required output for secondary control P2.

[0018] Although it has been stated above that the main control required output P1 is ensured, the output of the power storage device 201 actually used for main control only needs to be within the range of the rated output PA and may exceed the main control required output P1. The main control required output P1 is the minimum output of the power storage device 201 that can fully achieve the effect of main control. If the capacity of the power storage device 201 is equal to or greater than the main control required output P1, the power storage device 201 may output more than the main control required output P1 during main control. Furthermore, even if the estimation of the main control required output P1 by the power estimation device 401 is incorrect, the power storage device 201 may output more than the main control required output P1.

[0019] The primary control and secondary control may be executed simultaneously. In this case, as described above, the control surplus (PA-P1) becomes the upper limit of the output P2' of the secondary control power storage device 201.

[0020] <A-2. Power Inference Device> The power inference device 401 infers the required output for main control. In this specification, the output of the power storage device 201 includes at least one of the power charged from the power system 100 to the power storage device 201 and the power discharged from the power storage device 201 to the power system 100. Therefore, the required output for main control includes either or both of the power charged from the power system 100 to the power storage device 201 for main control and the power discharged from the power storage device 201 to the power system 100 for main control.

[0021] In the following description, stabilization control for stabilizing the power system 100 in which a disturbance has occurred is taken as an example of primary control. Normal control for the power system 100 in which no disturbance has occurred is taken as an example of secondary control. Normal control includes frequency control and load leveling control. Frequency control is control for charging and discharging so that the system frequency reaches a reference value. Load leveling control is control for suppressing load fluctuations by charging the power storage device 201 during time periods when the load is light and discharging from the power storage device 201 during time periods when the load is heavy.

[0022] The output of the power storage device 201 required for stabilization control is referred to as the “required stabilization output.” The power inference device 401 may infer the required stabilization output at a future time, or may infer the required stabilization output at a past time or at the present time.

[0023] Fig. 3 is a block diagram showing the configuration of the power inference device according to embodiment 1. In Fig. 3, the power inference device 401 is configured to include an acquisition unit 41, a power system simulation unit 42, a learning unit 43, an inference unit 44, an output unit 45, and a recording unit 46.

[0024] The acquisition unit 41 acquires system information representing the state of the power system 100 (hereinafter also referred to as the "system state"). The system information includes information representing the system configuration, such as the on / off state of switches and topology. The system information also includes information on the power sources and equipment connected to the power system 100, such as the rated capacity of the generators 111 and 112, the rated capacity of the inverter power supplies 121 and 122, the rated capacity of the power storage device 201, and the type of transmission line through which each power flows. The system information also includes measurement values ​​measured by the measurement devices 131-136, such as node voltages, phases, branch currents, generator outputs, and load amounts. The acquisition unit 41 also acquires system simulation conditions (hereinafter referred to as the "simulation conditions"). The simulation conditions include, for example, the assumed disturbance type, the location of the disturbance occurrence, the simulation target date and time, and the increment in the simulation of the upper output limit of the power storage device 201. Note that the information acquired by the acquisition unit 41 is not limited to the above example.

[0025] The acquiring unit 41 may acquire the power grid information by a user such as an operator of the power inference device 401 inputting the power grid information into the power inference device 401. The power inference device 401 may also be configured to cooperate with an external system connected to the communication network 170, such as an EMS (supply and demand control system), a SCADA (supervisory control and data acquisition system), a DAS (power distribution automation system), or a digital twin system, and the acquiring unit 41 may acquire the power grid information by automatically inputting the power grid information from these systems.

[0026] The system simulation unit 42 performs a power flow calculation and a detailed stability calculation using the system information and simulation conditions stored in the recording unit 46, and stores the results (simulation results) in the recording unit 46. Details of the system simulation will be described later.

[0027] The learning unit 43 creates an inference model using the system information stored in the recording unit 46 and the simulation results of the system simulation unit 42, and stores the model in the recording unit 46. In other words, the learning unit 43 functions as a creation unit that creates an inference model. The inference model is a model of the relationship between the system state and the stabilization required output (kW) of the power storage device 201 required for that system state.

[0028] The inference unit 44 infers the required stabilized output (kW) based on the inference model and system information stored in the recording unit 46 , and stores the result in the recording unit 46 .

[0029] The output unit 45 determines the output (kW) of the power storage device 201 that should be secured for stabilization control based on the required stabilization output (kW) inferred by the inference unit 44 and transmits this to the power control device 301.

[0030] The recording unit 46 stores the system information acquired by the acquisition unit 41, the simulation conditions, the simulation results by the system simulation unit 42, the inference model created by the learning unit 43, the inference results by the inference unit 44, etc. The recording unit 46 is configured by a database constructed on, for example, a PC (Personal Computer).

[0031] The system simulation unit 42 uses the system information acquired from the recording unit 46 to perform a power flow calculation and a detailed stability calculation assuming that the system is stabilized by the power storage device 201. At this time, the system simulation unit 42 does not change the system information, but generates multiple candidates by changing only the output of the power storage device 201 that can be used for stabilization control (hereinafter referred to as "stabilization usable output"), and stores the detailed stability calculation results for the multiple candidates in the recording unit 46.

[0032] The system information used in the system simulation includes the output (active power and reactive power) of the generators 111, 112 or the inverter power supplies 121, 122, node voltage, frequency, line flow (active power and reactive power, or active current and reactive current), etc.

[0033] The detailed stability calculation may assume the occurrence of a single system disturbance or multiple system disturbances occurring simultaneously or consecutively, such as a power source failure, a load failure, a short circuit, or a ground fault.

[0034] The power system simulation unit 42 may use PSLF or PSS / E, or the L method or Y method developed by the Central Research Institute of the Electric Power Industry, as an analysis program for performing power flow calculations and detailed stability calculations. Note that any other method, program, or tool may also be used for power system simulation.

[0035] The system simulation unit 42 performs detailed stability calculations for multiple candidates and determines whether the power system 100 is stable or unstable. Examples of evaluation indices for determining stability or instability include, but are not limited to, frequency, rate of change of frequency (RoCoF), generator phase angle, and generator output. For example, the system simulation unit 42 may determine that the power system 100 is stable if the phase difference between the center of inertia phase angle and each power source is within a threshold value. The system simulation unit 42 may also determine that the power system 100 is stable if the rate of change of frequency (RoCoF) or the lowest frequency point (Nadir) is within an appropriate range. Furthermore, evaluation indices newly devised or specified by the user may be used.

[0036] The system simulation unit 42 obtains the minimum stabilization usable output (kW) that can stabilize the power system 100 for the same system information, and determines this as the required stabilization output (kW).

[0037] 4 to 7 show example results of detailed stability calculations for four candidates. A stabilization result of OK means that the system was stabilized after stabilization control was performed, and a stabilization result of NG means that the system was unstable even after stabilization control was performed. For candidate 1 shown in FIG. 4, the stabilized usable output power PB is 10 kW, and the stabilization result is NG. For candidate 2 shown in FIG. 5, the stabilized usable output power PB is 20 kW, and the stabilization result is NG. For candidate 3 shown in FIG. 6, the stabilized usable output power PB is 30 kW, and the stabilization result is OK. For candidate 4 shown in FIG. 7, the stabilized usable output power PB is 40 kW, and the stabilization result is OK. From these results, the system simulation unit 42 determines that 30 kW, the minimum stabilized usable output power (kW) that could stabilize the power system 100 in the simulation, is the required stabilized output power (kW).

[0038] The learning unit 43 uses the system information and system simulation results acquired from the recording unit 46 to generate an inference model, which is a statistical model for inferring the required stabilized output (kW) for the target system information (system state), using a machine learning method. An example of the machine learning method is a supervised machine learning method, but is not limited to this. For example, the learning unit 43 may use an unsupervised machine learning method or reinforcement learning. Furthermore, the learning unit 43 may use a statistical method, an optimization method, or the like.

[0039] Examples of supervised machine learning techniques include, but are not limited to, decision trees, regression trees, random forests, random forest regression, support vector machines, support vector regression, neural networks, deep learning, etc. The learning unit 43 may use either a classification technique or a regression technique.

[0040] The inference model has two input / output patterns, a first input / output pattern and a second input / output pattern, and the processing by the learning unit 43 and the inference unit 44 differs for each input / output pattern, so these processes will be explained using Figures 8 and 9.

[0041] 8 shows the processing of the learning unit 43 that creates an inference model 504 for the first input / output pattern, and the processing of the inference unit 44 that performs inference using the inference model 504 for the first input / output pattern. Fig. 9 shows the processing of the learning unit 43 that creates an inference model 515 for the second input / output pattern, and the processing of the inference unit 44 that performs inference using the inference model 515 for the second input / output pattern.

[0042] The inference model 504 of the first input / output pattern is a model that outputs a required output power (kW) for stabilization in response to given system information. In this case, the learning unit 43 receives system information 501 and the like as feature quantities and receives a required output power (kW) for stabilization as an answer. Based on these, the learning unit 43 learns the answer to the feature quantities and creates the inference model 504. The learning unit 43 may perform learning using a machine learning regression method, and the output of the inference model 504 may be a continuous value. Alternatively, the learning unit 43 may perform learning using a machine learning classification method, and the output of the inference model 504 may be a discrete value. In this case, for example, category 0 is 0 kW, category 1 is 100 kW, etc.

[0043] The inference model 515 of the second input / output pattern is a model that outputs a stable / unstable flag indicating whether the power system 100 is stable or unstable after a disturbance occurs, for given system information and stabilization available output (kW). The stable / unstable flag is, for example, assumed to be 1 indicating stability and 0 indicating instability. In this case, the learning unit 43 receives the system information 511 as the first feature, the stabilization available output (kW) 512 as the second feature, and the stable / unstable flag 513 as the answer to the first feature and the second feature. Based on these, the learning unit 43 learns the answers to the first feature and the second feature and creates the inference model 515.

[0044] The inference unit 44 uses the inference model created by the learning unit 43 to infer the required stabilized output (kW) under the conditions of the measured system information or the system information stored in the recording unit 46 .

[0045] In the case of the inference model 504 of the first input / output pattern shown in FIG. 8 , the inference unit 44 inputs system information 505 to the inference model 504, obtains a required stabilization output (kW) 506 for the system state represented by the system information 505, and outputs this to the output unit 45. As shown in FIG. 8 , the inference model 504 may be provided for each expected disturbance type, such as an inference model 504A for a ground fault, an inference model 504B for a power supply drop, and a model 504C for a load drop. Furthermore, the inference model 504 may be provided for each disturbance case defined as a combination of a disturbance type and a disturbance occurrence point or location, or may be provided separately for charging and discharging. In this case, the inference unit 44 aggregates the required stabilization output (kW) obtained by each inference model 504, calculates the maximum value, and outputs this to the output unit 45 as the final required stabilization output (kW) 506. When inference models are provided separately for charging and discharging, the inference unit 44 calculates the maximum value for each charging or discharging and outputs this to the output unit 45 as the final required stabilized output (kW) 506.

[0046] In the case of the inference model 515 of the second input / output pattern shown in FIG. 9 , the inference unit 44 inputs fixed system information 516 to the inference model 515, while discretely changing the stabilized available power output (kW) and inputting the result to the inference model 515, thereby performing multiple inferences. The inference unit 44 then outputs the minimum stabilized available power output (kW) at which the stable / unstable flag becomes stable to the output unit 45 as the required stabilized power output (kW) 520. As shown in FIG. 9 , the inference model 515 may be provided for each expected disturbance type, each disturbance case, and each charging or discharging state, as in the case of FIG. 8 , such as an inference model 515A for a ground fault, an inference model 515B for a power source drop, and a model 515C for a load drop. In this case, the inference unit 44 aggregates the required stabilized power output (kW) obtained by each inference model 515 and outputs this to the output unit 45 as the final required stabilized power output (kW) 506.

[0047] The output unit 45 outputs the required stabilized power output (kW) inferred by the inference unit 44 to the power control device 301. The required stabilized power output (kW) output to the power control device 301 may be the required stabilized power output (kW) inferred by the inference unit 44 plus a margin. This margin is basically a positive value to provide a margin for stabilization control, but it may also be a negative value.

[0048] The method of output from the power inference device 401 to the power control device 301 may be transmission, screen display, or other method. Furthermore, the information output from the power inference device 401 to the power control device 301 is not limited to the required stabilized output (kW), but may also include the output of the power storage device 201 that can be used for normal control (hereinafter referred to as "normally usable output"), the system information used for inference, and the results of a system simulation used for learning that has conditions similar to those of the system information used for inference.

[0049] 3, the learning unit 43 and the inference unit 44 may be configured as separate devices. For example, the acquisition unit 41, the system simulation unit 42, the learning unit 43, and the recording unit 46 may form an inference model creation device that creates an inference model. Alternatively, the power inference device may be configured by an acquisition unit that acquires the inference model created by the inference model creation device, and an inference unit 44 that uses the inference model to infer the main control required output at the time of inference.

[0050] <A-3. Power control device> Fig. 10 is a block diagram showing the configuration of the power control device 301. As shown in Fig. 10, the power control device 301 is configured to include an acquisition unit 31, a normal output determination unit 32, a control command determination unit 33, an output unit 34, and a recording unit 35.

[0051] The acquisition unit 31 acquires measurement information, equipment information, control parameter information, and the output of the power inference device 401. The measurement information is measurement information obtained by the measurement devices 131-136, and includes, for example, information such as the phase angle (deg), voltage (V or pu), frequency (Hz), or current (A) of the measurement point. The equipment information includes, for example, the rated capacity (kW) of the storage battery PCS. The control parameter information includes, for example, the grid reference frequency (Hz). The output of the power inference device 401 includes, for example, the required output for stabilization (kW). Note that the information acquired by the acquisition unit 31 described above is only an example. The measurement information, equipment information, and control parameter information acquired by the acquisition unit 31 change depending on the control logic of normal control or stabilization control.

[0052] The normal-time output determining unit 32 determines the normal-time available output based on the required stabilized output (kW) acquired by the acquiring unit 31, and stores the determined output in the recording unit 35. Specifically, the normal-time available output is the remainder obtained by subtracting the required stabilized output from the maximum output of the power storage device 201. Alternatively, the normal-time available output may be the remainder to which a positive or negative value has been added. Note that if information on the normal-time available output is output from the power inference device 401 to the power control device 301, the normal-time output determining unit 32 is not necessary.

[0053] The control command determiner 33 generates commands for normal operation control and stabilization control based on the upper output limit value for normal operation control determined by the normal operation output determiner 32, and outputs the commands to the output unit 34. The normal operation control may be, for example, but is not limited to, a method of controlling the frequency deviation, which is the difference between the measured frequency and the grid reference frequency, so as to reduce the frequency deviation. The stabilization control may be, for example, but is not limited to, a method of controlling the synchronous generator by simulating its behavior based on the measured value.

[0054] The output unit 34 outputs the control command determined by the control command determination unit 33 to the power storage device 201 .

[0055] The recording unit 35 stores the information acquired by the acquisition unit 31 and the upper limit of the output power under normal control determined by the normal power output determination unit 32 .

[0056] <A-4. Flowchart> Fig. 11 is a flowchart showing the process of creating an inference model for the first input / output pattern by the power inference device 401. The process of creating an inference model will be described below with reference to the flow of Fig. 11 .

[0057] First, in step S101, the acquisition unit 41 acquires measurement information (actual values) at multiple inference points. This measurement information corresponds to system information. Then, in step S102, the system simulation unit 42 reflects the measurement information acquired in step S101 in a simulation model. Then, in step S103, the system simulation unit 42 reflects simulation conditions for system disturbances in the simulation model.

[0058] Next, in step S104, the power system simulation unit 42 sets the stabilized usable power output to an initial value. The initial value of the stabilized usable power output is, for example, 0 kW. Alternatively, the initial value of the stabilized usable power output may be specified by the user on each occasion.

[0059] Thereafter, in step S105, the system simulation unit 42 executes a power flow calculation and a detailed stability calculation. Here, the system simulation unit 42 uses a program capable of performing an effective value analysis.

[0060] Next, in step S106, the system simulation unit 42 determines whether the power system 100 is stable or unstable based on the detailed stability calculation result.

[0061] Thereafter, in step S107, the system simulation unit 42 determines whether the stabilized usable output is equal to or less than the maximum output of the power storage device 201. If the stabilized usable output is less than the maximum output of the power storage device 201 in step S107, the system simulation unit 42 increases the stabilized usable output in step S108 and returns to the processing of step S105. In this way, the system simulation unit 42 changes the stabilized usable output from the initial value to the maximum output to simulate the state of the power system.

[0062] If the stabilized usable output is equal to or greater than the maximum output of the power storage device 201 in step S107, the system simulation unit 42 calculates the minimum value of the stabilized usable output when the power system becomes stable in the simulation in step S109, and sets this as the required stabilized output.

[0063] Next, in step S110, the system simulation unit 42 determines whether or not calculations have been performed for all disturbance cases. If it is determined in step S110 that there are disturbance cases that have not yet been calculated, the processing of the system simulation unit 42 returns to step S102.

[0064] If calculations have been completed for all disturbance cases in step S110, the power inference device 401 determines whether or not processing has been completed for all time points in step S111. If there are any unprocessed time points remaining, the processing of the power inference device 401 returns to step S102.

[0065] If the processing has been completed for all time points in step S111, in step S112 the learning unit 43 creates an inference model based on the measurement information acquired in step S101 and the required output for stabilization control calculated in step S109. The learning unit 43 may create an inference model for each type of system disturbance, or may create one inference model that includes all types of system disturbance.

[0066] Then, in step S113, the learning unit 43 stores the inference model in the recording unit 46.

[0067] In the case of the process of creating an inference model for the second input / output pattern, the process of step S107 is not required. Then, in step S111, the learning unit 43 creates an inference model based on the measurement information acquired in step S101, the stabilized usable output (kW) set in step S104 or step S108, and the stable or unstable result (stable / unstable flag) determined in step S106.

[0068] 12 is a flowchart showing the inference process performed by the power inference device 401. The inference process will be described below along the flow of FIG.

[0069] First, in step S201, the inference unit 44 acquires measurement information (actual values) at the time of inference. This measurement information corresponds to system information. Next, in step S202, the inference unit 44 inputs the measurement information acquired in step S401 into an inference model and acquires the required stabilization output as an inferred value.

[0070] Note that this is an example of an inference model for the first input / output pattern. In the case of an inference model for the second input / output pattern, the inference unit 44 inputs the measurement information and the stabilized usable output (kW) to the inference model. At this time, the stabilized usable output (kW) is input as a discrete value ranging from 0 to the maximum output of the storage battery. The inference unit 44 then obtains the stable or unstable result as the output of the inference model, and obtains the minimum value of the stabilized usable output (kW) when stable as the inferred value of the stabilized output.

[0071] Next, in step S203, the output unit 45 outputs to the power control device 301 the storage battery capacity required for stabilization control obtained in step S202.

[0072] Thereafter, in step S204, the inference unit 44 determines whether it is the inference update period. If it is the update period, the processing of the inference unit 44 returns to step S201. If it is not the update period, the inference unit 44 determines in step S205 whether the inference has ended. If the inference has not ended, the inference unit 44 returns to the processing of step S204, and if the inference has ended, the power inference device 401 ends the inference processing.

[0073] 13 is a flowchart showing the processing of the power control device 301. The processing of the power control device 301 will be described below along the flow of FIG.

[0074] First, in step S301 , the acquisition unit 31 acquires the required stabilized power output (kW) from the power inference device 401 .

[0075] Next, in step S302, the normal-state output determining unit 32 sets the normal-state usable output (kW) based on the required stabilized output (kW) acquired in step S301.

[0076] Then, in step S303, the control command determination unit 33 determines a control command to cause the power storage device 201 to perform normal control within the range of the normal usable output (kW) set in step S302, and normal control is carried out.

[0077] Next, in step S304, the power control device 301 determines whether a grid disturbance has occurred. If a grid disturbance has occurred, in step S305, the control command determination unit 33 determines a control command for causing the power storage device 201 to perform stabilization control within the range of the maximum output that the power storage device 201 can output at that moment, and stabilization control is performed.

[0078] Thereafter, in step S306, the power control device 301 determines whether to end the stabilization control. If the stabilization control is not to be ended, the processing of the power control device 301 returns to step S305. If the stabilization control is to be ended, the processing of the power control device 301 proceeds to step S307.

[0079] If no grid disturbance has occurred in step S304, or if stabilization control is to be terminated in step S306, the power control device 301 determines in step S307 whether or not to terminate control of the power storage device 201. If control of the power storage device 201 is not to be terminated in step S307, the processing of the power control device 301 returns to step S303. If control of the power storage device 201 is to be terminated in step S307, the processing of the power control device 301 ends.

[0080] The power inference device 401 according to the first embodiment infers a required main control output, which is the charging / discharging power required for the main control of a power storage device that charges / discharges power to / from a power grid, by combining the main control with a secondary control that is implemented with a lower priority than the main control based on a control purpose different from that of the main control. The power inference device 401 includes an acquisition unit 41, a system simulation unit 42, a learning unit 43, and an inference unit 44. The acquisition unit 41 acquires system information representing the state of the power grid 100. The system simulation unit 42 calculates the required main control output corresponding to the acquired system information by simulating the state of the power grid 100 based on the acquired system information. The learning unit 43 generates an inference model of a first input / output pattern that can infer the required main control output from any system information based on multiple pairs of corresponding system information and required main control output. The inference unit 44 inputs the system information at the time of inference into the inference model of the first input / output pattern to infer the required main control output at the time of inference. The power storage device 201 determines the remaining output, for example, obtained by subtracting the required output for primary control estimated from the rated output, as the charge / discharge power in secondary control.

[0081] The power inference device 401 may also infer the main control required power using an inference model of a second output pattern. In this case, the system simulation unit 42 simulates whether main control can be achieved based on the system information and the main control usable output. The learning unit 43 then creates an inference model of the second output pattern that can infer whether main control can be achieved from the system information and the main control usable output based on the system information, the main control usable output, and whether main control can be achieved. The inference unit 44 inputs the system information and the main control usable output at the time of inference into the inference model of the second output pattern to infer whether main control can be achieved at the time of inference, and infers the minimum value of the main control usable output that can achieve main control as the main control required output at the time of inference.

[0082] <B. Second Embodiment> As shown in Fig. 2, the configuration of a power inference device 402 according to a second embodiment is the same as the configuration of the power inference device 401 according to the first embodiment. The operation of the power system simulation unit 42 of the power inference device 402 is different from that of the power inference device 401, and will be described below.

[0083] In the first embodiment, the system simulation unit 42 gradually increased the stabilized usable output (kW) used in the system simulation from the initial value to the maximum output of the storage battery, and performed a simulation of the detailed stability calculation for each stabilized usable output (kW).

[0084] In contrast to this, in the second embodiment, the system simulation unit 42 performs a simulation of the detailed stability calculation by setting the stabilized available output (kW) to an extremely large value, for example, a value equal to or greater than the rated output PA of the power storage device 201. Then, as shown in Fig. 14 , the system simulation unit 42 obtains the maximum value PN of the output (kW) of the power storage device 201 used for stabilization control in the detailed stability calculation, and determines this as the stabilized output (kW).

[0085] According to this method, the required stabilized output (kW) for the grid state is obtained by a single simulation, but it may exceed the rated output (maximum output) of the power storage device 201. Therefore, the required stabilized output (kW) inferred by the inference model created using the simulation results may also exceed the rated output of the power storage device 201. In that case, the output unit 45 may reduce the inferred required stabilized output to the rated output of the power storage device 201 and output it to the power control device 301.

[0086] Fig. 15 is a flowchart showing the process of creating an inference model by the power inference device 402. Compared to the flow of Fig. 11 described in the first embodiment, the flow of Fig. 15 includes step S401 instead of step S104, and step S402 instead of steps S106 to S109.

[0087] In step S401 , the system simulation unit 42 sets the stabilized available output (kW) to a value equal to or greater than the rated output of the power storage device 201 .

[0088] In step S402, the system simulation unit 42 calculates the required stabilization output (kW) from the result of the detailed stability calculation.

[0089] In the power inference device 402 according to the second embodiment, the power system simulation unit 42 simulates the state of the power system when the main control is performed by setting the main control available output to an arbitrary value equal to or greater than the maximum output of the power storage device 201, and calculates the maximum value of the charge / discharge power of the power storage device 201 in the simulation as the main control required output. In this way, the main control required output can be obtained in one simulation.

[0090] <C. Third Embodiment> Fig. 16 is a block diagram showing the configuration of a power inference device 403 according to a third embodiment. The power inference device 403 includes a system stabilization control calculation unit 47 in addition to the configuration of the power inference devices 401 and 402 according to the first and second embodiments. Differences from the first and second embodiments will be described below.

[0091] The system simulation unit 42 reflects the system state in a simulation model and performs power flow calculations and detailed stability calculations in consideration of power source restrictions, load restrictions, and stabilization control by the power storage device 201. For a certain system state, the system simulation unit 42 determines the amount of power source restrictions and load restrictions required for system stabilization and the required output for stabilization by optimization calculations such as metaheuristics.

[0092] The learning unit 43 receives input of grid information as a feature and input of the required stabilized output (kW) as an answer. Based on these inputs, the learning unit 43 creates an inference model that outputs the required stabilized output for the grid information. The output of the inference model may include the amount and location of power supply restriction and the amount and location of load restriction.

[0093] The inference unit 44 infers the output required for stabilization for the grid information using the inference model created by the learning unit 43. When the output of the inference model includes the amount and location of power supply restriction and the amount and location of load restriction, the inference result of the inference unit 44 may include the amount and location of power supply restriction and the amount and location of load restriction for the grid information.

[0094] The grid stabilization control calculation unit 47 calculates the amount and location of power supply restriction and the amount and location of load restriction based on the inference result of the inference unit 44. The grid stabilization control calculation unit 47 can determine the amount and location of power supply restriction and the amount and location of load restriction by judging the inference result of the inference unit 44 and repeatedly performing detailed stability calculations. When the inference result of the inference unit 44 includes the amount and location of power supply restriction and the amount and location of load restriction for the grid information, the grid stabilization control calculation unit 47 may adopt the amount and location of power supply restriction and the amount and location of load restriction inferred by the inference unit 44.

[0095] The output unit 45 transmits the required stabilized output (kW) to the power control device 301, and also transmits information on the amount and location of power supply restriction and the amount and location of load restriction calculated by the system stabilization control calculation unit 47 to the control devices 141-148 via the communication network 170. In response to this, the control devices 141-148 control the loads 101-104, generators 111, 112, inverter power supplies 121, 122, etc., and execute power supply restriction and load restriction.

[0096] The power system stabilization control calculation unit 47 may be provided as a power system stabilization device outside the power inference device 403. In this case, the power system stabilization device and the power inference device 403 transmit and receive data to and from each other via the communication network 170. That is, the power system stabilization system of the third embodiment includes the power inference device 403 and a power system stabilization device that calculates the amount and location of power supply restriction and load restriction of the power system 100 at the time of inference based on the required stabilized output inferred by the power inference device 403.

[0097] Fig. 17 is a flowchart showing the process of creating an inference model by the power inference device 403. Compared to the flow of Fig. 11 described in the first embodiment, the flow of Fig. 17 includes step S501 instead of steps S104 to S109.

[0098] In step S501, the system simulation unit 42 performs a power flow calculation and a detailed stability calculation, and calculates the amount and location of power supply restrictions required for system stabilization, the amount and location of load restrictions, and the required stabilization output (kW) of the power storage device 201.

[0099] Fig. 18 is a flowchart showing the inference process by the power inference device 403. Compared to the flow in Fig. 12 described in the first embodiment, the flow in Fig. 18 includes step S601 added between step S203 and step S204.

[0100] In step S601, the inference unit 44 determines the amount and location of power supply restriction and the amount and location of load restriction based on the required stabilized output obtained in step S202. The determined content is then transmitted from the output unit 45 via the communication network 170 to the control devices 141-148, which then issue commands to restrict the power supply and the load.

[0101] <D. Hardware Configuration> The above-described power inference devices 401-403 and power control device 301 are realized by a processing circuit 81 shown in FIG. 19. That is, the processing circuit 81 includes the components of the power inference devices 401-403 and the power control device 301. The processing circuit 81 may be implemented by dedicated hardware, or may be implemented by a processor that executes programs stored in memory. Examples of the processor include a central processing unit, a processing unit, an arithmetic unit, a microprocessor, a microcomputer, and a DSP (Digital Signal Processor).

[0102] When the processing circuit 81 is dedicated hardware, the processing circuit 81 may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof. The functions of each unit of the power inference devices 401-403 and the power control device 301 may be realized by multiple processing circuits 81, or the functions of each unit may be realized together by a single processing circuit.

[0103] When the processing circuit 81 is a processor, the functions of the power inference devices 401-403 and the power control device 301 are realized by a combination of software, etc. (software, firmware, or software and firmware). The software, etc. is written as a program and stored in memory. As shown in FIG. 20 , the processor 82 applied to the processing circuit 81 realizes the functions of the respective units by reading and executing a program stored in memory 83. That is, the power inference devices 401-403 and the power control device 301 include memory 83 for storing a program that, when executed by the processing circuit 81, results in the processing of the respective units of the power inference devices 401-403 and the power control device 301. In other words, the program causes a computer to execute the procedures or methods of the respective units of the power inference devices 401-403 and the power control device 301. Here, the memory 83 may be, for example, a non-volatile or volatile semiconductor memory such as a RAM (Random Access Memory), a ROM (Read Only Memory), a flash memory, an EPROM (Erasable Programmable Read Only Memory), or an EEPROM (Electrically Erasable Programmable Read Only Memory), a HDD (Hard Disk Drive), a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, a DVD (Digital Versatile Disk) and its drive device, or any storage medium that will be used in the future.

[0104] The above describes a configuration in which the functions of the power inference devices 401-403 and the power control device 301 are realized by either hardware or software, etc. However, the present invention is not limited to this, and a configuration in which some of the functions of the power inference devices 401-403 and the power control device 301 are realized by dedicated hardware and other parts are realized by software, etc.

[0105] As described above, the processing circuit can realize each of the above-described functions by hardware, software, or a combination of these.

[0106] It should be noted that the embodiments can be freely combined, and each embodiment can be modified or omitted as appropriate. The above description is an example in all respects. It is understood that countless variations not illustrated can be envisioned.

[0107] 31 Acquisition unit, 32 Normal output determination unit, 33 Control command determination unit, 34 Output unit, 35 Recording unit, 41 Acquisition unit, 42 System simulation unit, 43 Learning unit, 44 Inference unit, 45 Output unit, 46 Recording unit, 47 System stabilization control calculation unit, 81 Processing circuit, 82 Processor, 83 Memory, 100 Power system, 101-104 Load, 111, 112 Generator, 121, 122 Inverter power supply, 131-136 Measuring device, 141-148 Control device, 151, 152 Transformer, 170 Communication network, 201 Power storage device, 301 Power control device, 401-403 Power inference device, 450 Power control system.

Claims

1. A power inference device that infers the main control required output, which is the charge / discharge power required for the main control of a power storage device that performs charging and discharging with a power grid, by combining a main control and a sub-control that is performed with a lower priority than the main control based on a different control purpose, An acquisition unit that acquires system information representing the status of the power system, A system simulation unit calculates the main control required output corresponding to the acquired system information by simulating the state of the power system based on the acquired system information, A creation unit creates an inference model capable of inferring the main control required output from any system information based on a plurality of sets of corresponding system information and main control required outputs, The inference model includes an inference unit that inputs the system information at the time of inference and infers the main control required output at the time of inference, Based on the remainder obtained by subtracting the inferred main control required output from the maximum output, the output of the power storage device in the sub-control is determined. Power inference device.

2. The aforementioned main control is a power system stabilization control performed in the event of a disturbance in the power system. The aforementioned sub-control is a control that is performed at least during normal operation of the power system. The power inference device according to claim 1.

3. The creation unit creates the inference model using machine learning. The power inference device according to claim 1 or claim 2.

4. The system simulation unit calculates the required output for main control by simulating the state of the power system when the main control usable output, which is the charge / discharge power of the power storage device usable for main control, is changed from an initial value to a maximum output. The power inference device according to claim 1 or claim 2.

5. The system simulation unit simulates the state of the power system when the main control is performed with the main control available output, which is the charge / discharge power of the power storage device that can be used for the main control, set to an arbitrary value greater than or equal to the maximum output, and calculates the maximum value of the charge / discharge power of the power storage device in the simulation as the main control required output. The power inference device according to claim 1 or claim 2.

6. A power inference device that infers the main control required output, which is the charge / discharge power required for the main control of a power storage device that charges from or discharges to a power system, by combining a main control and a sub-control that is performed with a lower priority than the main control for a control purpose different from the main control, An acquisition unit that acquires system information representing the status of the power system, A system simulation unit that simulates whether the main control can be achieved based on the acquired system information and the main control usable output, which is the charge / discharge power of the power storage device usable for the main control, A creation unit creates an inference model that can infer whether the main control can be achieved from the system information and the main control available output, based on the system information, the main control available output, and whether the main control can be achieved. The inference model includes an inference unit that inputs the system information and the main control available output at the time of inference to infer whether the main control can be achieved at the time of inference, and infers the minimum value of the main control available output that makes the main control achievable as the main control required output at the time of inference, The remainder obtained by subtracting the inferred main control required output from the maximum output becomes the output of the power storage device in the sub-control. Power inference device.

7. The inference unit calculates the amount and location of the power supply limit and load limit of the power system at the time of the inference, based on the inferred main control required output. The power inference device according to claim 1 or claim 6.

8. A power inference device according to claim 1 or claim 6, The power control device comprises: determining the main control required output inferred by the power inference device as the charge / discharge power of the power storage device for main control, and determining the remainder obtained by subtracting the main control required output from the maximum output of the power storage device as the charge / discharge power of the power storage device available for sub-control. Power control system.

9. A power inference device according to claim 1 or claim 6, A power system stabilization device that calculates the amount and location of power supply and load limits for the power system at the time of inference, based on the inferred main control required output, System stabilization system.

10. An inference model creation device for creating an inference model for inferring the main control required output, which is the charge / discharge power required for the main control of a power storage device that performs charging and discharging with a power grid by combining a main control and a sub-control that is performed with a lower priority than the main control based on a different control purpose, An acquisition unit that acquires system information representing the status of the power system, A system simulation unit calculates the main control required output corresponding to the acquired system information by simulating the state of the power system based on the acquired system information, The system comprises: a creation unit that creates an inference model capable of inferring the main control required output from any system information based on a plurality of sets of corresponding system information and main control required outputs, An inference model generation device.

11. A power inference device that infers the main control required output, which is the charge / discharge power required for the main control of a power storage device that performs charging and discharging with a power grid, by combining a main control and a sub-control that is performed with a lower priority than the main control based on a different control purpose, An acquisition unit that acquires an inference model capable of inferring the main control required output from arbitrary system information, which is created based on a plurality of sets of system information representing the state of the power system and the main control required output, The inference model includes an inference unit that inputs the system information at the time of inference and infers the main control required output at the time of inference, Based on the remainder obtained by subtracting the inferred main control required output from the maximum output, the output of the power storage device in the sub-control is determined. Power inference device.

12. A power inference method for inferring the main control required output, which is the charge / discharge power required for the main control of a power storage device that charges from or discharges to a power system, by combining a main control and a sub-control that is performed with a lower priority than the main control for a control purpose different from that of the main control, System information representing the state of the power system is obtained, By simulating the state of the power system based on the acquired system information, the required main control output corresponding to the acquired system information is calculated. Based on multiple sets of corresponding system information and main control required outputs, an inference model is created that can infer the main control required output from any system information. The system information at the time of inference is input to the inference model to infer the main control required output at the time of inference. Methods for extrapolating electricity.