Information processing system, information processing method, and program

Abstract

An information processing system (10) comprises: an acquisition unit (14) that acquires a time waveform of the power consumption of a dwelling unit (101); an estimation unit (15) that estimates charging power consumption information indicating a time waveform of the power consumption of a charging system (20) of an electric vehicle (23) installed in the dwelling unit (101) on the basis of the acquired time waveform of the power consumption of the dwelling unit (101); and a first prediction unit (16) that selects a machine learning model on the basis of the estimated charging power consumption information and predicts the future power consumption of the charging system (20) using the selected machine learning model.

Description

Information processing system, information processing method, and program

[0001] The present disclosure relates to an information processing system, an information processing method, and a program.

[0002] Patent Document 1 discloses a charge / discharge control device that effectively uses surplus power generated by solar power, etc. The charge / discharge control device disclosed in Patent Document 1 charges or discharges a battery mounted on a vehicle.

[0003] JP 2010-233408 A

[0004] The present disclosure provides an information processing system and the like that can predict future power consumption of a charging system for electric vehicles installed in a facility based on the time waveform of power consumption at the facility.

[0005] An information processing system according to one aspect of the present disclosure includes an acquisition unit that acquires a time waveform of power consumption at a facility; an estimation unit that estimates charging power consumption information indicating a time waveform of power consumption of a charging system for an electric vehicle installed at the facility based on the acquired time waveform of power consumption at the facility; and a first prediction unit that selects a machine learning model based on the estimated charging power consumption information and predicts future power consumption of the charging system using the selected machine learning model.

[0006] An information processing system according to one aspect of the present disclosure includes an acquisition unit that acquires a power adjustment request that requests an increase or decrease in power demand in a target area during a future target period, and a control unit that sets, based on the acquired power adjustment request, a time period during which at least some of a plurality of charging systems installed in a plurality of facilities located in the target area will charge electric vehicles.

[0007] An information processing method according to one aspect of the present disclosure is an information processing method executed by a computer system, and includes the steps of acquiring a time waveform of power consumption at a facility, estimating charging power consumption information indicating a time waveform of power consumption of a charging system for an electric vehicle installed at the facility based on the acquired time waveform of power consumption at the facility, and selecting a machine learning model based on the estimated charging power consumption information and predicting future power consumption of the charging system using the selected machine learning model.

[0008] An information processing method according to one aspect of the present disclosure is an information processing method executed by a computer system, and includes the steps of acquiring a power adjustment request requesting an increase or decrease in power demand in a target area for a future target period, and setting, based on the acquired power adjustment request, a time period during which at least some of a plurality of charging systems installed in a plurality of facilities located in the target area will charge electric vehicles.

[0009] A program according to one aspect of the present disclosure is a program for causing the computer system to execute the information processing method.

[0010] The information processing system etc. disclosed herein can predict future power consumption of a charging system for electric vehicles installed in a facility from the time waveform of power consumption of the facility.

[0011] Fig. 1 is a block diagram showing the functional configuration of an information processing system according to an embodiment. Fig. 2 is a flowchart showing a specific example of an operation for predicting power consumption of an information processing system according to an embodiment. Fig. 3 is a diagram showing a method for predicting power consumption of a charging system. Fig. 4 is a flowchart showing a specific example of an operation for adjusting power demand of an information processing system according to an embodiment. Fig. 5 is a diagram for explaining a shift in a time period during which an electric vehicle is charged.

[0012] Hereinafter, the embodiments will be described in detail with reference to the drawings. Note that the embodiments described below are all comprehensive or specific examples. The numerical values, shapes, materials, components, component placement and connection forms, steps, and step order shown in the following embodiments are merely examples and are not intended to limit the present disclosure. Furthermore, among the components in the following embodiments, components not recited in independent claims will be described as optional components.

[0013] It should be noted that the drawings are schematic diagrams and are not necessarily strict illustrations. In addition, in the drawings, substantially the same components are denoted by the same reference numerals, and overlapping descriptions may be omitted or simplified.

[0014] (Embodiment) [Configuration] First, the configuration of an information processing system according to an embodiment will be described. Fig. 1 is a block diagram showing the functional configuration of an information processing system according to an embodiment.

[0015] The information processing system 10 is a system that can predict future power consumption of multiple charging systems 20 installed in multiple dwelling units 101 located in a target area 100. The dwelling units 101 are an example of a facility. The target area 100 refers to an area that belongs to the same power grid (such as an area under the jurisdiction of a power company). In addition to the information processing system 10, FIG. 1 illustrates multiple charging systems 20, multiple power metering devices 30, a power management system 50, and a weather information management system 60. These systems and devices will be described below.

[0016] The information processing system 10 is a system realized by, for example, one or more server devices. Specifically, the information processing system 10 includes a communication unit 11, an information processing unit 12, and a storage unit 13.

[0017] The communication unit 11 is a communication module (communication circuit) that enables the information processing system 10 to communicate with a plurality of charging systems 20, a power management system 50, a weather information management system 60, and the like via a wide area communication network 90 such as the Internet. The communication performed by the communication unit 11 is, for example, wired communication, but may also be wireless communication. There is no particular limitation on the communication standard used for the communication.

[0018] The information processing unit 12 performs information processing to predict future power consumption of the multiple charging systems 20. The information processing unit 12 is realized, for example, by a microcontroller, but may also be realized by a processor. The information processing unit 12 includes, as functional components, an acquisition unit 14, an estimation unit 15, a first prediction unit 16, a second prediction unit 17, a creation unit 18, and a control unit 19. The functions of the acquisition unit 14, the estimation unit 15, the first prediction unit 16, the second prediction unit 17, the creation unit 18, and the control unit 19 are realized, for example, by the microcontroller or processor constituting the information processing unit 12 executing a computer program stored in the storage unit 13.

[0019] The storage unit 13 is a storage device that stores information necessary for the information processing, computer programs executed by the information processing unit 12, etc. The storage unit 13 is realized by, for example, a hard disk drive (HDD), but may also be realized by a semiconductor memory or the like.

[0020] The charging system 20 is provided in each of a plurality of dwelling units 101 located in the target area 100, and is a system for charging electric vehicles in the dwelling units 101. Specifically, the charging system 20 includes a communication device 21, a charging device 22, and an electric vehicle 23.

[0021] The communication device 21 is, for example, a wireless router, and includes a communication module (communication circuit) that enables the charging system 20 to communicate with the information processing system 10 via the wide area communication network 90. ​​The communication device 21, for example, performs wireless communication with the charging device 22 and connects to the wide area communication network 90 by wired communication. The communication standard for communication performed by the communication device 21 is not particularly limited.

[0022] The charging device 22 charges the electric vehicle 23 (more specifically, the secondary battery 24 provided in the electric vehicle 23) using AC power obtained from the power source. Specifically, the charging device 22 is realized by a circuit that outputs AC power suitable for charging the secondary battery 24 provided in the electric vehicle 23 to the electric vehicle 23, a microcontroller for charging control, and the like.

[0023] The electric vehicle 23 is a vehicle used by a user residing in the dwelling unit 101. The electric vehicle 23 is, for example, an EV (Electric Vehicle) that runs by driving a motor, but may also be a vehicle that runs using both an engine and a motor, such as a PHV (Plug-in Hybrid Vehicle) or a PHEV (Plug-in Hybrid EV). The electric vehicle 23 includes a secondary battery 24 as an energy source for running the electric vehicle 23.

[0024] The secondary battery 24 is a secondary battery (storage battery) mounted on the electric vehicle 23, and is charged by the charging device 22 while the vehicle is still mounted on the electric vehicle 23. Specifically, the secondary battery 24 is a lithium ion battery or the like. Note that the electric vehicle 23 is, for example, a four-wheeled vehicle, but may also be a two-wheeled vehicle or a three-wheeled vehicle.

[0025] The power measurement device 30 is provided in each of the multiple dwelling units 101 and measures the time waveform of power consumption in the dwelling units 101. The power measurement device 30 also communicates with the power management system 50 via the wide area communication network 90, thereby transmitting the measured time waveform of power consumption (time series data of power consumption) to the power management system 50. The power measurement device 30 is, for example, a smart meter, but may also be a distribution board with a power measurement function.

[0026] The power management system 50 is a system realized by one or more server devices. The power management system 50 receives power data from multiple power measurement devices 30, and thereby manages the time waveform of power consumption (power consumption amount) in each of multiple dwelling units 101. The power management system 50 is used, for example, by an electric power company (i.e., a business operator that owns the power grid) that supplies power to multiple dwelling units 101 located in the target area 100.

[0027] The weather information management system 60 is a system realized by one or more server devices. The weather information management system 60 provides weather forecast information and past weather information to the information processing system 10 by communicating with the information processing system 10 via a wide area communication network 90. ​​The weather forecast information and the weather information each include solar radiation information, temperature information, etc.

[0028] [Power Consumption Prediction Operation] In order to calculate the electricity bill for each of the multiple dwelling units 101, the power consumption of the entire dwelling unit 101 is measured by the power metering device 30. However, it is difficult to extract (estimate) the power consumption of a specific device installed in the dwelling unit 101 from the power consumption of the entire dwelling unit 101, and it is even more difficult to predict the future power consumption of the specific device.

[0029] In response to this, the information processing system 10 can predict future power consumption of the charging system 20 installed in the dwelling unit 101 based on the power consumption of the entire dwelling unit 101. A specific example of the power consumption prediction operation of the information processing system 10 will be described below. FIG. 2 is a flowchart of a specific example of the power consumption prediction operation of the information processing system 10. An example will be described below in which the time waveform of the power consumption of the charging system 20 on the first target day (future) is predicted in the evening of the day before the first target day (for example, around 3:00 PM to 4:00 PM).

[0030] The acquisition unit 14 acquires temperature information (weather forecast information) for the target area 100 on a first target day and temperature information for the target area 100 on a past second target day (such as the day before the first target day) by communicating with the weather information management system 60 using the communication unit 11 (S11). More specifically, the temperature information is information indicating the predicted maximum temperature and the predicted minimum temperature on the first target day or a past second target day.

[0031] The acquisition unit 14 also acquires, from the storage unit 13, baseline data that has been stored in advance in the storage unit 13 (S12). The baseline data includes actual data on the time waveform of power consumption in each of multiple dwelling units 101 located in the target area 100. The actual data on the time waveform of power consumption in a certain dwelling unit 101 is data measured in the past by a power measurement device 30 installed in that dwelling unit 101, and indicates the time waveform of the overall power consumption of the certain dwelling unit 101 (in other words, the time waveform of power consumption in the main circuit of the certain dwelling unit 101). The actual data on the time waveform of power consumption included in the baseline data is actual data covering a relatively long period of time, for example, several months to several years.

[0032] The baseline data is provided to the information processing system 10 from the power management system 50 (electric power company), for example, by the acquisition unit 14 communicating with the power management system 50 using the communication unit 11, but may be provided via any route.

[0033] Next, the first prediction unit 16 predicts the time waveform of power consumption on the first target day for each of the multiple dwelling units 101 located in the target area 100 based on the temperature information for the first target day acquired in step S11 and the baseline data acquired in step S12 (S13). The time waveform of power consumption here means the time waveform of the total power consumption of the dwelling units 101.

[0034] In the baseline data, the actual data of the time waveform of power consumption is linked to the month, date, day of the week, whether it is a holiday, time, and temperature information. In step S12, the first prediction unit 16, for example, refers to the baseline data to identify a day (approximate day) whose day of the week, holiday, and temperature information (temperature information acquired in step S11) are most similar to the first target day. The first prediction unit 16 can predict the time waveform of power consumption of each of the multiple dwelling units 101 on the identified approximate day as the time waveform of power consumption of each of the multiple dwelling units 101 on the first target day.

[0035] Next, the estimation unit 15 estimates the amount of power consumption of each of the multiple charging systems 20 corresponding to the multiple dwelling units 101 (S14). Specifically, the estimation unit 15 estimates the amount of power consumption of each of the multiple charging systems 20 using a machine learning model stored in the storage unit 13.

[0036] This machine learning model is a trained model that uses as training data performance data for a dwelling unit that associates the following information (1) to (5) for a given day. The training data may be performance data obtained for a dwelling unit 101 located in the target area 100, or performance data obtained for a dwelling unit located in an area other than the target area 100. Note that if a solar power generation system is not installed in the dwelling unit 101, the information (5) is omitted.

[0037] (1) Time waveform of the power consumption of the charging system on that day. (2) Time waveform of the power consumption of the entire dwelling unit on that day. (3) Temperature information on that day. (4) Date information on that day (month, day, day of the week, whether it is a holiday or not). (5) Time waveform of the power generated by the solar power generation system on that day.

[0038] The estimation unit 15 inputs into the machine learning model the time waveform of power consumption of the dwelling unit 101 on the past second target day, which is part of the baseline data acquired in step S12, and the temperature information on the past second target day and the date information on the past second target day acquired in step S11, thereby being able to acquire the time waveform of power consumption on the past second target day of the charging system 20 installed in the dwelling unit 101. The estimation unit 15 can estimate the amount of power consumption of the charging system 20 installed in the dwelling unit 101 from the acquired time waveform.

[0039] Hereinafter, information indicating the amount of power consumption of the charging system 20 will also be referred to as charging power consumption information. The processing of step S14 can be rephrased as processing of estimating charging power consumption information for each of the multiple charging systems 20 based on the time waveforms of the power consumption of the multiple dwelling units 101 acquired in step S13. Note that the charging power consumption information may be information indicating the time waveforms of the power consumption of the charging system 20.

[0040] More specifically, three types of machine learning models are stored in the memory unit 13, and in step S14, for example, any one of the three types (for example, machine learning model M) is used.

[0041] Here, three types of machine learning models will be described. For example, the above-described learning data is divided into three types, namely, learning data L, learning data M, and learning data S, according to the magnitude (L (large), M (medium), S (small)) of the power consumption of the charging system 20 included in the learning data. This allows three types of machine learning models, namely, a machine learning model L trained on the learning data L, a machine learning model M trained on the learning data M, and a machine learning model S trained on the learning data S, to be generated and stored in the storage unit 13. The three types of machine learning models are constructed based on mutually different learning data.

[0042] The first prediction unit 16 selects a machine learning model for each of the plurality of charging systems 20 based on the charging power consumption information of each of the plurality of charging systems 20 estimated in step S14 (S15).

[0043] Specifically, the first prediction unit 16 selects one of three types of machine learning models for each of the multiple charging systems 20. The first prediction unit 16 selects machine learning model L for a charging system 20 whose power consumption amount estimated in step S14 is large. The first prediction unit 16 selects machine learning model M for a charging system 20 whose power consumption amount estimated in step S14 is standard. The first prediction unit 16 selects machine learning model S for a charging system 20 whose power consumption amount estimated in step S14 is small. The power consumption amount is determined using a threshold, for example. The selection of the machine learning model may be performed based on the total power consumption of the dwelling unit 101.

[0044] Next, the first prediction unit 16 predicts the power consumption on the first target day of each of the multiple charging systems 20 corresponding to the multiple dwelling units 101 (S16). FIG. 3 is a diagram showing a method for predicting the power consumption of the charging system 20.

[0045] FIG. 3A shows the time waveform of the power consumption of the entire dwelling unit 101 predicted in step S13. The first prediction unit 16 inputs the time waveform of the power consumption of the dwelling unit 101 on the first target day predicted in step S13 (corresponding to FIG. 3A), the temperature information for the first target day acquired in step S12, and the date information for the first target day into the machine learning model selected in step S15. This allows the first prediction unit 16 to predict the power consumption (time waveform of power consumption) of the charging system 20 installed in the dwelling unit 101 on the first target day. FIG. 3B shows the time waveform of the power consumption of the charging system 20 predicted by the machine learning model in step S16. Note that in FIG. 3B, the actual measured value (correct data) of the time waveform of the power consumption of the charging system 20 is indicated by a dotted line.

[0046] In this way, the first prediction unit 16 selectively uses three types of machine learning models to predict the time waveform of the power consumption of the charging system 20 on the first target day, based on the magnitude (estimated value) of the past power consumption of the charging system 20. This makes it possible to improve the prediction accuracy compared to predicting the time waveform of the power consumption of the charging system 20 on the first target day by fixedly using one type of machine learning model. Note that two types of machine learning models may be prepared, or four or more types may be prepared.

[0047] To further improve the prediction accuracy, the first prediction unit 16 may correct the time waveform of the power consumption of the charging system 20 predicted by the machine learning model using a predetermined algorithm. (c) of Fig. 3 shows the time waveform of the power consumption of the charging system 20 corrected using the predetermined algorithm. (c) of Fig. 3 shows the actual measured value (correct data) of the time waveform of the power consumption of the charging system 20 using a dotted line.

[0048] The time waveform of the power consumption of the charging system 20 has a characteristic shape (top hat shape), and therefore, the prediction accuracy can be improved by correcting (shaping the waveform) using a predetermined algorithm.

[0049] After the process of step S16, the second prediction unit 17 predicts the power consumption resulting from the operation of the charging system 20 (charging of the electric vehicle 23) on the first target day in the target area 100 (which can also be considered the entire target area 100) based on the power consumption on the first target day of each of the multiple charging systems 20 predicted in step S16 (S17). The second prediction unit 17 can predict the time waveform of the power consumption resulting from the operation of the charging system 20 on the first target day in the target area 100 by adding together the time waveforms of the power consumption on the first target day of each of the multiple charging systems 20 predicted in step S16.

[0050] As described above, the information processing system 10 can predict the future power consumption of each of the multiple charging systems 20 installed in the multiple dwelling units 101 based on the time waveforms of the power consumption of the multiple dwelling units 101.

[0051] [Power Demand Adjustment Operation] In recent years, photovoltaic power generation systems have become increasingly common, and there are cases in which an excess supply of power occurs during the daytime hours when the photovoltaic power generation systems generate power in the target area 100. Here, each of the multiple charging systems 20 is normally set to charge the electric vehicles 23 required for one day during the late night hours.

[0052] Therefore, the information processing system 10 controls some of the charging systems 20 of the multiple charging systems 20 so that the charging of the electric vehicles 23 by the some of the charging systems 20 is performed during the daytime. In other words, the information processing system 10 shifts (sets) the time periods during which the some of the multiple charging systems 20 charge the electric vehicles 23 from late night hours to daytime hours, thereby increasing the demand for electricity during the daytime hours in the target area 100 and stabilizing the power system (reducing excess power supply).

[0053] A specific example of such an operation for adjusting the power demand will be described below. Fig. 4 is a flowchart of a specific example of the operation for adjusting the power demand of the information processing system 10. The following describes an example in which the time waveform of the power consumption of the charging system 20 on the first target day (future) is predicted in the evening of the day before the first target day (for example, between 3:00 PM and 4:00 PM).

[0054] The acquisition unit 14 of the information processing system 10 acquires a power adjustment request from the power management system 50 (electric power company) by communicating with the power management system 50 using the communication unit 11 (S21). The power adjustment request is information requesting an increase in power demand in the target area 100 during a target period on a first target date. The power adjustment request specifies the target period and the amount of power that needs to be increased during the target period.

[0055] The first prediction unit 16 predicts the power consumption on the first target day of each of the multiple charging systems 20 corresponding to the multiple dwelling units 101 located in the target area 100 (S22). The second prediction unit 17 predicts the power consumption resulting from the operation of the charging systems 20 (charging of the electric vehicles 23) on the first target day in the target area 100 (S23). The processing of steps S22 and S23 is similar to the power consumption prediction operation described above, and therefore detailed description thereof will be omitted.

[0056] The creation unit 18 creates a shift plan for shifting the time periods during which some of the multiple charging systems 20 charge the electric vehicles 23 based on the acquired power adjustment request (S24). In other words, the process of creating a shift plan is a process of determining the charging systems 20 that will shift the time periods during which the electric vehicles 23 are charged. Fig. 5 is a diagram for explaining the shifting of the time periods during which the electric vehicles 23 are charged.

[0057] The bar graph shown in (a) of Figure 5 shows the predicted results (prediction results of the second prediction unit 17 in step S23) of the power consumption due to the operation of the charging system 20 (charging of the electric vehicle 23) on the first target day in the target area 100.

[0058] 5A or more during the daytime target period, the creation unit 18 shifts the time period during which some of the multiple charging systems 20 charge the electric vehicles 23 to the target period. At this time, because the power consumption of each of the multiple charging systems 20 on the first target day is predicted in step S22, the creation unit 18 can simulate how much power consumption will shift to the target period when it is assumed that the time period during which a certain charging system 20 charges the electric vehicles 23 is shifted from the late night hours to the daytime target period.

[0059] The creation unit 18 selects some of the multiple charging systems 20 so that the power consumption (power demand) during the target period exceeds a threshold, and creates a shift plan that shifts the time periods during which the selected some of the charging systems 20 charge the electric vehicles 23 during the target period. As a result, as shown in (b) of Fig. 5, the power demand during the target period can be increased to or above the threshold.

[0060] Below, a method for determining some of the multiple charging systems 20 (hereinafter also referred to as target systems) for which the time period for charging the electric vehicles 23 is to be shifted will be described. The multiple dwelling units 101 located in the target area 100 are divided into n groups, group 1 to group n (n is a natural number of 2 or greater). In other words, the multiple charging systems 20 are divided into n groups.

[0061] The control unit 19 selects, for example, m candidate groups (m is a natural number equal to or greater than 1) that are a part of the n groups, based on the priorities set for the n groups.

[0062] The creation unit 18 performs a simulation to determine whether the goal of increasing the power demand in the target period to a threshold or more will be achieved (whether the result will be as shown in FIG. 5(b) ) if the time periods during which the electric vehicles 23 are charged by the selected m candidate groups are shifted to the target period. The extent to which the power consumption can be shifted can be estimated using the power consumption of the multiple charging systems 20 on the first target day predicted in step S22.

[0063] If the simulation results show that the target has been achieved, the creation unit 18 confirms m candidate groups and determines the charging systems 20 belonging to the confirmed candidate groups (hereinafter also referred to as target groups) as target systems. If the simulation results show that the target has not been achieved, for example, the time periods during which the electric vehicles 23 are charged are not shifted, or the candidate groups are changed (added, deleted, or replaced) and the simulation is performed again.

[0064] In determining the target system, in order to prevent the same group from being determined as the target group repeatedly, once a group has been determined as the target group, its priority for selection as one of the m candidate groups is lowered. In other words, once a group has been determined as the target group, it becomes less likely to be determined as the target group for a certain period of time.

[0065] Note that this method of determining the target system (target group) is one example. The information processing system 10 may employ other algorithms that can determine the target system so that the goal is achieved.

[0066] After the shift plan is generated in step S24 in this way, the acquisition unit 14 acquires the generated shift plan (S25), and the control unit 19 transmits a control command to the target system indicated by the shift plan by communicating with the target system using the communication unit 11 based on the acquired shift plan (S26). The control command specifies a first time slot set by the control unit 19, which is at least a part of a time slot of a target period during which the electric vehicle 23 should be charged.

[0067] The charging device 22 of the target system receives the control command via the communication device 21 and charges the electric vehicle 23 during the first time slot specified in the received control command. That is, the charging device 22 of the target system shifts the time slot for charging the electric vehicle 23 from the late night time slot to the first time slot (at least part of the target period) based on the received control command. In this way, the control unit 19 sets the time slots during which at least some of the multiple charging systems 20 charge the electric vehicle 23 to overlap with at least part of the target period based on the acquired power adjustment request.

[0068] The charging devices 22 of the charging systems 20 other than the target system that did not receive the control command charge the electric vehicles 23 during the night hours as usual based on the past operation history information.

[0069] As described above, the information processing system 10 can increase the power demand during the target period (daytime hours) in the target area 100 by shifting the time periods during which some of the multiple charging systems 20 charge the electric vehicles 23 to the target period based on the power adjustment request.

[0070] Based on the power adjustment request, the information processing system 10 can also reduce (suppress) the power demand for a target period in the target area 100. For example, if the power adjustment request requests a reduction in power demand during late-night hours (target period), the information processing system 10 can reduce the power demand for the target period in the target area 100 by shifting the time periods during which some of the multiple charging systems 20 charge the electric vehicles 23 to time periods other than the target period based on the power adjustment request.

[0071] [Modification] The plurality of processes described in the operation of adjusting power demand in the above embodiment may be shared and performed by the information processing system 10 and another system.

[0072] For example, the processes of steps S22 to S24 may be performed by the information processing system 10 (first information processing system), and the processes of steps S21, S25, and S26 may be performed by the second information processing system. In this case, when the second information processing system acquires (receives) a power adjustment request from the power management system 50, it requests the first information processing system to create a shift plan. When the second information processing system acquires (receives) a shift plan from the first information processing system in response to the request, it controls some of the multiple charging systems 20 based on the acquired shift plan. Specifically, the second information processing system shifts the time period during which some of the multiple charging systems 20 charge the electric vehicles 23 to a time period outside the target period.

[0073] Furthermore, for example, the processing of steps S22 to S23 may be performed by the information processing system 10 (first information processing system), and the processing of steps S21 and S24 to S26 may be performed by the second information processing system. That is, when the second information processing system acquires (receives) a power adjustment request from the power management system 50, it requests the first information processing system to predict the power consumption of the multiple charging systems 20. When the second information processing system acquires (receives) prediction results from the first information processing system in response to the request, it creates a shift plan based on the prediction results and controls some of the multiple charging systems 20 based on the created shift plan. Specifically, the second information processing system shifts the time period during which some of the multiple charging systems 20 charge the electric vehicles 23 to a time period outside the target period.

[0074] Furthermore, in the above embodiment, it has been described that the machine learning model for predicting the time waveform of the power consumption of the charging system 20 is selected based on the magnitude of the past power consumption of the charging system 20. However, the machine learning model may be selected based on a trend (parameter) obtained from the time waveform of the past power consumption of the charging system 20, and it is not essential that the machine learning model be selected based on the magnitude of the past power consumption. Furthermore, one machine learning model may be used in a fixed manner, and it is not essential that a machine learning model be selected.

[0075] Furthermore, in the above embodiment, an example has been described in which the time waveform of the power consumption of the charging system 20 installed in the dwelling unit 101 is predicted. However, the information processing system 10 may also predict the time waveform of the power consumption of the charging system 20 installed in a facility other than the dwelling unit 101 that is located in the target area 100.

[0076] Similarly, in the above embodiment, an example has been described in which the time period for charging the electric vehicle 23 performed by the charging system 20 installed in the dwelling unit 101 is shifted in order to adjust the demand for electricity. However, the information processing system 10 may also shift the time period for charging the electric vehicle 23 performed by the charging system 20 installed in a facility other than the dwelling unit 101 in order to adjust the demand for electricity. Examples of facilities other than the dwelling unit 101 include office buildings, factories, research facilities, commercial facilities, nursing homes, medical facilities, and accommodation facilities. The facilities correspond to consumers from the perspective of the electric power company.

[0077] [Effects, etc.] Hereinafter, examples of techniques that can be obtained from the disclosure of this specification will be given, and effects, etc. that can be obtained from the exemplified techniques will be described.

[0078] Technology 1 is an information processing system 10 including an acquisition unit 14 that acquires a time waveform of power consumption of a dwelling unit 101, an estimation unit 15 that estimates charging power consumption information indicating a time waveform of power consumption of a charging system 20 of an electric vehicle 23 installed in the dwelling unit 101 based on the acquired time waveform of power consumption of the dwelling unit 101, and a first prediction unit 16 that selects a machine learning model based on the estimated charging power consumption information and predicts future power consumption of the charging system 20 using the selected machine learning model. The dwelling unit 101 is an example of a facility.

[0079] Such an information processing system 10 can predict future power consumption of the charging system 20 installed in the dwelling unit 101 from the time waveform of power consumption of the dwelling unit 101.

[0080] Technology 2 is the information processing system 10 of Technology 1, in which an acquisition unit 14 acquires time waveforms of power consumption of a plurality of dwelling units 101, including a dwelling unit 101, located in a target area 100, and a plurality of charging systems 20, including a charging system 20, are installed in the plurality of dwelling units 101, an estimation unit 15 estimates charging power consumption information of each of the plurality of charging systems 20 based on the acquired time waveforms of power consumption of the plurality of dwelling units 101, and a first prediction unit 16 selects a machine learning model for each of the plurality of charging systems 20 based on the estimated charging power consumption information of each of the plurality of charging systems 20, and predicts future power consumption of each of the plurality of charging systems 20 using the machine learning model selected for the charging system 20.

[0081] Such an information processing system 10 can predict future power consumption of a plurality of charging systems 20 installed in a plurality of dwelling units 101 from the time waveform of power consumption of the plurality of dwelling units 101 .

[0082] Technology 3 is an information processing system 10 of Technology 1 or 2, in which the machine learning model is selected from at least three machine learning models constructed based on mutually different learning data, and each of the at least three machine learning models takes as input the acquired time waveform of power consumption of the dwelling unit 101 and outputs the time waveform of power consumption of the charging system 20.

[0083] Such an information processing system 10 can improve the accuracy of predicting future power consumption of the charging system 20 by selecting and using one machine learning model from at least three machine learning models based on estimated charging power consumption information.

[0084] Technology 4 is the information processing system 10 of Technology 2, further including a second prediction unit 17 that predicts the power consumption resulting from the operation of the charging systems 20 in the target area 100 in the future based on the predicted future power consumption of each of the multiple charging systems 20.

[0085] Such an information processing system 10 can predict the power consumption resulting from the operation of the charging system 20 in the target area 100 in the future.

[0086] Technology 5 is the information processing system 10 of Technology 4, in which the acquisition unit 14 further acquires a shift plan for a time period during which the multiple charging systems 20 charge the electric vehicles 23, the first prediction unit 16 predicts future power consumption of each of the multiple charging systems 20 in a case where the acquired shift plan is realized, using a machine learning model selected for the charging system 20, and the second prediction unit 17 predicts power consumption resulting from operation of the charging system 20 in the future in the target area 100 in a case where the acquired shift plan is realized, based on the predicted future power consumption of each of the multiple charging systems 20.

[0087] Such an information processing system 10 can predict the power consumption resulting from the operation of the charging system 20 in the target area 100 in the future based on the shift plan.

[0088] Technology 6 is an information processing system 10 including an acquisition unit 14 that acquires a power adjustment request that requests an increase or decrease in the power demand in a target area 100 for a future target period, and a control unit 19 that sets, based on the acquired power adjustment request, a time period during which at least some of a plurality of charging systems 20 installed in a plurality of dwelling units 101 located in the target area 100 will charge electric vehicles 23.

[0089] The information processing system 10 can adjust the demand for electricity by shifting the time period during which the charging system 20 charges the electric vehicle 23 .

[0090] Technology 7 is the information processing system 10 of Technology 6, in which the acquisition unit 14 acquires time waveforms of power consumption of multiple dwelling units 101 located in the target area 100, and the information processing system 10 further includes an estimation unit 15 that estimates charging power consumption information indicating the time waveforms of power consumption of each of multiple charging systems 20 based on the acquired time waveforms of power consumption of the multiple dwelling units 101.

[0091] The information processing system 10 can adjust the demand for electricity by shifting the time periods during which the multiple charging systems 20 charge the electric vehicles 23 based on the estimated charging power consumption information.

[0092] Technology 8 is the information processing system 10 of Technology 6 or 7, further including a creation unit 18 that creates a shift plan for time periods during which the multiple charging systems 20 charge the electric vehicles 23 based on the acquired power adjustment request, and a control unit 19 that sets time periods during which at least some of the multiple charging systems 20 charge the electric vehicles 23 based on the created shift plan.

[0093] The information processing system 10 can adjust the demand for electricity by shifting the time periods during which the multiple charging systems 20 charge the electric vehicles 23 based on the created shift plan.

[0094] Technique 9 is the information processing system 10 of any of Techniques 6 to 8, in which the acquired power adjustment request requests an increase in the power demand in the target area 100 during a target period, and the control unit 19 sets, based on the acquired power adjustment request, the time periods during which at least some of the multiple charging systems 20 charge the electric vehicles 23 to overlap with at least part of the target period.

[0095] Such an information processing system 10 can increase the demand for electricity during the target period by setting the time periods during which at least some of the multiple charging systems 20 charge the electric vehicles 23 to overlap with at least part of the target period.

[0096] Technique 10 is an information processing method executed by a computer system, and includes step S13 of acquiring a time waveform of power consumption of the dwelling unit 101, step S14 of estimating charging power consumption information indicating a time waveform of power consumption of a charging system 20 of an electric vehicle 23 installed in the dwelling unit 101 based on the acquired time waveform of power consumption of the dwelling unit 101, and step S16 of selecting a machine learning model based on the estimated charging power consumption information and using the selected machine learning model to predict future power consumption of the charging system 20. The computer system is, for example, information processing system 10.

[0097] Such an information processing method makes it possible to predict future power consumption of the charging system 20 installed in the dwelling unit 101 from the time waveform of power consumption of the dwelling unit 101 .

[0098] Technique 11 is an information processing method executed by a computer system, and includes: a step S21 of acquiring a power adjustment request that requests an increase or decrease in the power demand in a target area 100 for a future target period; and a step S26 of setting, based on the acquired power adjustment request, a time period during which at least some of the multiple charging systems 20 installed in multiple dwelling units 101 located in the target area 100 will charge electric vehicles 23.

[0099] Such an information processing method can adjust the demand for electricity by shifting the time period during which the charging system 20 charges the electric vehicle 23 .

[0100] Technique 12 is a program for causing a computer system to execute the information processing method of Technique 10 or 11.

[0101] According to such a program, the computer system can predict future power consumption of the charging system 20 installed in the target area 100 or adjust the power demand in the target area 100.

[0102] (Other Embodiments) Although the embodiments have been described above, the present disclosure is not limited to the above-described embodiments.

[0103] For example, in the above-described embodiments, the information processing system may be realized by multiple devices or may be realized as a single device. Thus, the system in this specification may be configured by a single device or may be configured by multiple devices. When the system is realized by multiple devices, the components of the system may be distributed among the multiple devices in any manner.

[0104] Furthermore, the method of communication between the devices in the above-described embodiment is not particularly limited. Furthermore, a relay device (such as a broadband router, not shown) may be involved in the communication between the devices.

[0105] In the above-described embodiment, the processing performed by a specific processing unit may be performed by another processing unit. The order of multiple processing operations may be changed, or multiple processing operations may be performed in parallel.

[0106] In the above-described embodiments, each component may be realized by executing a software program suitable for that component, or by a program execution unit such as a CPU or processor reading and executing a software program recorded on a recording medium such as a hard disk or semiconductor memory.

[0107] Furthermore, each component may be realized by hardware. For example, each component may be a circuit (or integrated circuit). These circuits may form a single circuit as a whole, or each may be a separate circuit. Furthermore, each of these circuits may be a general-purpose circuit or a dedicated circuit.

[0108] Furthermore, the general or specific aspects of the present disclosure may be realized as a system, an apparatus, a method, an integrated circuit, a computer program, or a computer-readable recording medium such as a CD-ROM, or as any combination of a system, an apparatus, a method, an integrated circuit, a computer program, and a recording medium.

[0109] For example, the present disclosure may be realized as an information processing method executed by a computer system such as the information processing system of the above-described embodiment, or as a program (computer program product) for causing a computer system to execute the information processing method. Furthermore, the present disclosure may be realized as a computer-readable non-transitory recording medium on which such a program is recorded.

[0110] The present disclosure may be realized as a control method for a charging system executed by a computer system such as the charging system of the above-described embodiment, or as a program (computer program product) for causing a computer system to execute the control method. Furthermore, the present disclosure may be realized as a computer-readable non-transitory recording medium on which such a program is recorded.

[0111] In addition, this disclosure also includes forms obtained by applying various modifications to each embodiment that a person skilled in the art would think of, or forms realized by arbitrarily combining the components and functions of each embodiment within the scope that does not deviate from the intent of this disclosure.

[0112] REFERENCE SIGNS LIST 10 Information processing system 11 Communication unit 12 Information processing unit 13 Memory unit 14 Acquisition unit 15 Estimation unit 16 First prediction unit 17 Second prediction unit 18 Creation unit 19 Control unit 20 Charging system 21 Communication device 22 Charging device 23 Electric vehicle 24 Secondary battery 30 Power measurement device 50 Power management system 60 Weather information management system 90 Wide area communication network 100 Target area 101 Dwelling unit

Claims

1. An information processing system comprising: an acquisition unit that acquires a time waveform of power consumption of a facility; an estimation unit that estimates charging power consumption information indicating a time waveform of power consumption of a charging system of an electric vehicle installed in the facility based on the acquired time waveform of power consumption of the facility; and a first prediction unit that selects a machine learning model based on the estimated charging power consumption information and predicts future power consumption of the charging system using the selected machine learning model.

2. The acquisition unit acquires time waveforms of power consumption of a plurality of facilities including the facility located in a target area, a plurality of charging systems including the charging system are installed in the plurality of facilities, the estimation unit estimates the charging power consumption information of each of the plurality of charging systems based on the acquired time waveforms of power consumption of the plurality of facilities, and the first prediction unit selects the machine learning model for each of the plurality of charging systems based on the estimated charging power consumption information of each of the plurality of charging systems, and predicts future power consumption of each of the plurality of charging systems using the selected machine learning model for the charging system. The information processing system according to claim 1.

3. The machine learning model is a machine learning model selected from at least three machine learning models constructed based on different learning data, and each of the at least three machine learning models outputs a time waveform of power consumption of the charging system with the acquired time waveform of power consumption of the facility as an input. The information processing system according to claim 1 or 2.

4. Further comprising a second prediction unit that predicts power consumption caused by the operation of the charging system in the future of the target area based on the predicted future power consumption of each of the plurality of charging systems. The information processing system according to claim 2.

5. The acquisition unit further acquires a shift plan for time periods during which the plurality of charging systems charge the electric vehicle, and the first prediction unit predicts future power consumption of each of the plurality of charging systems when the acquired shift plan is realized, using the machine learning model selected for the charging system, and the second prediction unit predicts future power consumption in the target area caused by the operation of the charging system in the future when the acquired shift plan is realized, based on the predicted future power consumption of each of the plurality of charging systems. The information processing system according to claim 4.

6. An information processing system comprising: an acquisition unit that acquires a power adjustment request for increasing or decreasing the power demand in a target area during a future target period; and a control unit that sets a time period during which at least some of the plurality of charging systems installed in a plurality of facilities located in the target area charge the electric vehicle, based on the acquired power adjustment request.

7. The acquisition unit acquires time waveforms of power consumption of a plurality of facilities located in the target area, and the information processing system further includes an estimation unit that estimates charging power consumption information indicating time waveforms of power consumption of each of the plurality of charging systems, based on the acquired time waveforms of power consumption of the plurality of facilities. The information processing system according to claim 6.

8. Further comprising a creation unit that creates a shift plan for a time period during which the plurality of charging systems charge the electric vehicle, based on the acquired power adjustment request, and the control unit sets a time period during which at least some of the plurality of charging systems charge the electric vehicle, based on the created shift plan. The information processing system according to claim 6.

9. The acquired power adjustment request requests to increase the power demand in the target area during the target period, and the control unit sets, based on the acquired power adjustment request, a time period during which at least some of the plurality of charging systems charge the electric vehicle to overlap at least a part of the target period. The information processing system according to any one of claims 6 to 8.

10. An information processing method executed by a computer system, the method including: obtaining a time waveform of power consumption of a facility; estimating charging power consumption information indicating a time waveform of power consumption of a charging system of an electric vehicle installed in the facility based on the obtained time waveform of power consumption of the facility; and selecting a machine learning model based on the estimated charging power consumption information and predicting future power consumption of the charging system using the selected machine learning model.

11. An information processing method executed by a computer system, the method including: obtaining a power adjustment request for increasing or decreasing power demand in a target area in a future target period; and setting a time period during which at least some of a plurality of charging systems installed in a plurality of facilities located in the target area perform charging of electric vehicles based on the obtained power adjustment request.

12. A program for causing the computer system to execute the information processing method according to claim 10 or 11.

Images