System and method for estimating charge / discharge efficiency of a storage battery
The system estimates storage battery efficiency using machine learning and historical data, addressing the lack of environmental information to enhance power transaction accuracy and management.
Patent Information
- Application Number
- JP2021138437
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-08-26
- Publication Date
- 2024-11-13
- Estimated Expiration
- 2041-08-26
AI Technical Summary
Existing technologies require environmental information, such as temperature and aging deterioration information, of storage batteries, which is often unknown or not disclosed, making it impossible to accurately correct charging/discharging performance and plan operations.
A system and method that estimates charging and discharging efficiency of storage batteries using a machine learning engine, training with teacher data correlating past temperatures or solar power generation with efficiency, and communicating via a network to control charging/discharging based on estimated efficiency.
Enables accurate estimation of charging and discharging efficiency without direct measurement of battery temperature, improving power transaction accuracy and efficiency management.
Smart Images

Figure 0007582920000001 
Figure 0007582920000002 
Figure 0007582920000003
Abstract
Description
[Technical field]
[0001] The present invention relates to a technique for controlling charging and discharging a storage battery. [Background technology]
[0002] There is a technology that allows the power company to control the charging and discharging of the users' storage batteries, rather than just unilaterally adjusting the supply and demand of the power amount. According to this technology, the power intermediary company constantly collects information on the power usage of each user's storage battery. For example, the power intermediary business instructs each storage battery of the user on the amount of power to be charged and discharged according to the amount of power for supply and demand adjustment. In response, each storage battery charges and discharges power according to the received amount of power to be charged and discharged. The power intermediary business adjusts the amount of power for supply and demand adjustment so that there is as little difference as possible between the amount of power for supply and demand adjustment and the total amount of power charged and discharged by all storage batteries.
[0003] Conventionally, there is a technology for a Virtual Power Plant (VPP) that controls electric vehicles and home storage batteries installed in a user's home (see, for example, Patent Document 1). According to this technology, charge / discharge performance information (e.g., charge / discharge efficiency) of an electric device is corrected based on environmental information of the electric device (e.g., temperature and aging deterioration information of the storage battery). Then, the charge / discharge operation of the electric device is controlled according to an operation plan based on the corrected charge / discharge performance information.
[0004] The same applicant and inventor as the present application have previously filed patent applications for a technology that uses location information of a user terminal to improve the accuracy of predicting future power consumption (see, for example, Patent Document 2). There is also a technology that uses an unspecified number of posted messages to predict future power consumption (see, for example, Patent Document 3). [Prior art documents] [Patent documents]
[0005] [Patent Document 1] JP 2020-156149 A [Patent Document 2] Patent No. 6714489 [Patent Document 3] Patent No. 6812320 Summary of the Invention [Problem to be solved by the invention]
[0006] However, the technology described in Patent Document 1 requires environmental information (temperature and information on deterioration over time) of the storage battery itself. In the case of a typical home storage battery, the temperature of the storage battery itself is often unknown or is not disclosed to anyone other than the manufacturer. If environmental information of the storage battery itself cannot be obtained, it is ultimately impossible to correct the charge / discharge performance information of the power device, and an accurate charge / discharge plan cannot be made. In reality, according to ECHONET Lite (Device Object Detailed Specification Release P) (registered trademark), the storage battery class does not include property information related to temperature.
[0007] Both of the techniques described in Patent Documents 2 and 3 are techniques for predicting the amount of power consumed by a storage battery, and do not take into account the charging and discharging efficiency of the storage battery.
[0008] Therefore, the present invention estimates the charge / discharge efficiency of a storage battery. Systems and methods The purpose is to provide. [Means for solving the problem]
[0009] According to the present invention, there is provided an estimation device for estimating charge / discharge efficiency of a storage battery. and a home equipment device having a storage battery and a charge / discharge control device, The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is In the training stage, training is performed using teacher data that associates past temperatures with charging and discharging efficiency, and in the estimation stage, a machine learning engine is used to input the temperature at the target time and estimate the charging and discharging efficiency. and transmitting the estimated charge / discharge efficiency to the home equipment. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. It is characterized by:
[0010] The present invention According to An estimation device for estimating the charge / discharge efficiency of a storage battery connected to a photovoltaic power generation device and a home equipment device having a storage battery and a charge / discharge control device, The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is In the training stage, the system is trained using teacher data that correlates past solar power generation and charging / discharging efficiency, and in the estimation stage, a machine learning engine is used to input the solar power generation at the target time and estimate the charging / discharging efficiency. and transmitting the estimated charge / discharge efficiency to the home equipment. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. It is characterized by:
[0011] The present invention According to An estimation device for estimating the charge / discharge efficiency of a storage battery connected to a photovoltaic power generation device and a home equipment device having a storage battery and a charge / discharge control device, The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is In the training stage, training is performed using teacher data that associates past temperatures, solar power generation, and charge / discharge efficiency, and in the estimation stage, a machine learning engine is used to input the temperature and solar power generation at the target time and estimate the charge / discharge efficiency. and transmitting the estimated charge / discharge efficiency to the home equipment. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. It is characterized by:
[0012] The present invention system According to another embodiment of the present invention, Regarding the charge / discharge efficiency, the ratio of the amount of change in the remaining power to the amount of charge power is defined as the charge efficiency, and the ratio of the amount of discharge power to the amount of change in the remaining power is defined as the discharge efficiency. Estimation device In the training stage, the machine learning engine separately trains using teacher data in which the charging efficiency and the discharging efficiency are associated with the objective variable, and in the estimation stage, estimates the charging efficiency and the discharging efficiency separately. It is also preferable.
[0013] The present invention system According to another embodiment of the present invention, The teacher data further includes time zones and / or weather information as explanatory variables. It is also preferable.
[0014] The present invention system According to another embodiment of the present invention, The teacher data is the explanatory variable, Status information indicating when charging / discharging is controlled / normal power consumption, and / or Charge / discharge power amount during control, or charge / discharge power amount during normal operation Further comprising: Estimation device The machine learning engine is In the training stage, training is performed separately for the charge / discharge efficiency during control and the charge / discharge efficiency during normal operation. In the estimation step, the charge / discharge efficiency during control and the charge / discharge efficiency during normal operation are estimated separately. It is also preferable.
[0015] The present invention system According to another embodiment of the present invention, The maximum charge amount or maximum discharge amount during charging / discharging is controlled as follows: Maximum charging capacity = min (rated charging capacity, (battery capacity - remaining battery capacity) / charging efficiency) Maximum discharge amount = min (rated discharge amount, remaining battery power amount x discharge efficiency) It is also preferable.
[0019] According to the present invention, there is provided an estimation device for estimating charge / discharge efficiency of a storage battery. and a home equipment device having a storage battery and a charge / discharge control device, and the home equipment device communicates with the storage battery and the charge / discharge control device via a network. A method for estimating The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is In the training phase, the system is trained using teacher data that associates past temperatures with charging and discharging efficiency, and in the estimation phase, the system inputs the temperature at the target time and creates a machine learning engine that estimates charging and discharging efficiency. and transmitting the estimated charge / discharge efficiency to the home equipment device. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. It is characterized by:
[0020] The present invention According to An estimation device for estimating the charge / discharge efficiency of a storage battery connected to a photovoltaic power generation device and a home equipment device having a storage battery and a charge / discharge control device, The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is In the training stage, the system is trained using teacher data that correlates past solar power generation and charging / discharging efficiency, and in the estimation stage, the system inputs the amount of solar power generation at the target time and estimates the charging / discharging efficiency. The estimated charge / discharge efficiency is transmitted to the home equipment device using the The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. It is characterized by: The present invention According to Estimation device for estimating charge / discharge efficiency of a storage battery connected to a solar power generation device and a home equipment device having a storage battery and a charge / discharge control device, The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is In the training stage, training is performed using teacher data that associates past temperatures, solar power generation, and charge / discharge efficiency, and in the estimation stage, a machine learning engine is used to input the temperature and solar power generation at the target time and estimate the charge / discharge efficiency. and transmitting the estimated charge / discharge efficiency to the home equipment. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. It is characterized by: Effect of the Invention
[0021] The present invention Systems and methods According to this, the charging and discharging efficiency of the storage battery can be estimated. [Brief description of the drawings]
[0022] [Figure 1] FIG. 1 is a diagram showing a system configuration based on energy trading. [Diagram 2] FIG. 2 is a diagram showing the configuration of a home appliance; [Diagram 3] FIG. 2 is a functional configuration diagram of an estimation device according to the present invention. [Figure 4] FIG. 2 is an illustration of the training phase of the machine learning engine of the present invention. [Diagram 5]FIG. 11 is an explanatory diagram showing how to determine the amount of power when purchasing and selling electricity, taking into account the charge / discharge efficiency. [Figure 6] FIG. 1 is an explanatory diagram showing extended data in teacher data of a machine learning engine. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0023] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0024] FIG. 1 is a diagram showing the configuration of a system based on energy trading.
[0025] As shown in FIG. 1, each of the customer's premises equipment 2 is connected to a power network and a communication network. The home equipment device 2 requests the estimation device 1 to trade an amount of power desired by a user via a communication network. In addition, the home equipment device 2 transmits or receives power for the established power trade via a power network managed by the power transmission and distribution equipment device 3. The power transmission and distribution equipment device 3 is operated and managed by a power transmission and distribution business operator.
[0026] The estimation device 1 of the present invention estimates the "charge and discharge efficiency" of a storage battery as a home equipment device 2. The lower the charge and discharge efficiency of a storage battery, the less power it can charge compared to the amount of power it receives, and the more power it must discharge compared to the amount of power it should transmit. In other words, the charge and discharge efficiency is an important performance index for each storage battery. The estimation device 1 is operated and managed by an electric utility company, and provides transactions for buying and selling electricity for users. The estimation device 1 completes an electricity transaction requested by an indoor equipment device 2 via a communication network. The indoor equipment device 2 is controlled so as to charge and discharge the amount of electricity of the completed transaction to and from the electricity transmission and distribution equipment device 3.
[0027] 1, a weather server 4 is disposed in the communication network. The weather server 4 collects and distributes "temperature" for each region. By accessing the weather server 4, the estimation device 1 can obtain temperature and weather information for the region where the user's home equipment device 2 is disposed.
[0028] FIG. 2 is a diagram showing the configuration of the home appliance.
[0029] The indoor equipment 2 includes a storage battery 20, a distribution board 212, a charge / discharge control device 211, a smart meter 221, a Home Energy Management System (HEMS) gateway 222, a home gateway 23, a user terminal 24, a thermometer 25, and a photovoltaic power generation device (PhotoVoltaic) 26.
[0030] [Storage battery 20] The storage battery 20 is capable of charging (current flow) and discharging (reverse current). It may be a stationary battery, or may be mounted on an electric vehicle parked at a user's home. The power of the storage battery 20 is consumed by home appliances in the user's home. The storage battery 20 may be one that is charged (purchased electricity) from a power grid, or one that is charged (self-generated electricity) from a solar power generation device 26. Of course, it may also be one that is charged from other devices that have a power generation capacity. In addition, surplus electricity from the storage battery 20 can be discharged (sold) to the power grid.
[0031] [Charge / discharge control device 211] The charge / discharge control device 211 is connected to the power grid, and is also connected to the storage battery 20 and a HEMS gateway 222. The charge / discharge control device 211 has a function of charging and discharging the storage battery 20 to the power grid via a distribution board 212 and a smart meter 221. The charge / discharge control device 211 controls the charge and discharge functions in response to control signals from the HEMS gateway 222. The charge / discharge control device 211 generally has a function of converting AC electricity for the power grid and DC electricity for the storage battery, and is also called a "power conditioner."
[0032] [Distribution Board 212] The distribution board 212 is an electrical facility that distributes the power received from the power grid via the smart meter 221 and the power received from the storage battery 20 via the charge / discharge control device 211 to in-home electrical devices. For safety reasons, a circuit breaker, a ground fault circuit interrupter, and the like are collectively attached to the distribution board 212. The distribution board 212 can also transmit the power received from the charge / discharge control device 211 to the power grid via the smart meter 221.
[0033] [Smart Meter 221] The smart meter 221 is a power meter that digitally measures the power transmitted and received between the power grid and the distribution board 212 , and outputs the measured power value to the HEMS gateway 222 .
[0034] [HEMS Gateway 222] The HEMS gateway 222 communicates with the estimation device 1 through a communication network via a home gateway 23. The HEMS gateway 222 can also communicate with a user terminal 24 that can be operated by a user via the home gateway 23.
[0035] Application software for power trading is installed in the HEMS gateway 222. This allows status information on the storage battery 20 and the charge / discharge control device 211 to be sequentially transmitted to the estimation device 1. Specifically, the HEMS gateway 222 may transmit and receive data to and from the estimation device 1 in the data format of the OpenADR protocol, and transmit data to and from the charge / discharge control device 211 in the data format of the ECHONET Lite (registered trademark) protocol.
[0036] The HEMS gateway 222 can also transmit information on the user's power transaction received from the user terminal 24 to the estimation device 1 and receive information on the established power transaction from the estimation device 1. Specifically, it establishes the amount of power purchased and the amount of power sold in the user's household. The HEMS gateway 222 can also control the charge / discharge control device 211 according to the established power transaction.
[0037] [Home Gateway 23] The home gateway 23 is connected to a communication network, and is also connected to the HEMS gateway 222 and the user terminal 24. The home gateway 23 relays communication between the HEMS gateway 222 and the estimation device 1.
[0038] [User terminal 24] The user terminal 24 is, for example, a smartphone, and can communicate with the HEMS gateway 222 via the home gateway 23. The user terminal 24 can also connect to a communication network via the home gateway 23 and transmit power trading information to the estimation device 1 based on a user operation. In the case of selling electricity, the user terminal 24 transmits to the estimation device 1 the discharge time (discharge time zone), the discharged energy amount, and the selling price as electricity trading information. In the case of purchasing electricity, the user terminal 24 transmits to the estimation device 1 the charging time (charging time zone), the charging capacity, and the purchasing price as electricity trading information. The user terminal 24 can then clearly display the power transaction information established by the estimation device 1 and the storage battery state information transmitted from the HEMS gateway 222 to the user.
[0039] [Thermometer 25] The thermometer 25 outputs the temperature value around the indoor equipment 2 to the HEMS gateway 222.
[0040] [Solar power generation equipment 26] The solar power generation device 26 generates electricity using sunlight, outputs the self-generated electricity to the storage battery 20, and also outputs the amount of generated electricity to the HEMS gateway 222. The amount of power generated by the solar power generation device 26 is proportional to the amount of solar radiation. The amount of solar radiation also affects the temperature. For this reason, the amount of power generated by the solar power generation device, which is proportional to the temperature, can be used.
[0041] According to FIG. 2, the HEMS gateway 222 of the present invention can transmit not only power information but also the temperature value of the thermometer 25 and the amount of power generated by the photovoltaic power generation device 26 to the estimation device 1. The thermometer 25 is not essential to the present invention as the estimation device 1, and it is sufficient if the temperature around the storage battery 20 can be acquired. The temperature around the home equipment device 2 can also be acquired from a weather server 4. The weather server 4 may transmit to the estimation device 1 not only past and current temperatures but also forecast temperatures for future time periods for each location (e.g., an identifier representing an area divided into a mesh shape).
[0042] FIG. 3 is a functional configuration diagram of the estimation device according to the present invention. FIG. 4 is an illustration of the training stage of the machine learning engine of the present invention.
[0043] 3, the estimation device 1 includes a machine learning engine 10, an environmental information receiving unit 11, and a charge / discharge efficiency transmitting unit 12. These functional components are realized by executing a program that causes a computer mounted on the device to function. The process flow of these functional components can also be understood as a method for estimating the charge / discharge efficiency of the device.
[0044] [Machine Learning Engine 10] The machine learning engine 10 is typically a regression model, for example, but may also apply random forests. Statistically, regression is the construction of a linear regression model, Y=f(X), between explanatory variable X and target variable Y. When there are multiple explanatory variables, it becomes multiple regression. Random forest is an ensemble learning algorithm that uses decision trees as weak learners, and uses a large number of decision trees trained using randomly sampled training data.
[0045] The machine learning engine 10 has, for example, the following three patterns of embodiment. First pattern: Estimate "charge / discharge efficiency" according to "temperature" Second pattern: Estimating "charge / discharge efficiency" according to "amount of solar power generation" Third pattern: Estimate "charge / discharge efficiency" according to "temperature and solar power generation amount"
[0046] <First pattern> (Training stage) Training is performed using teacher data that associates past temperatures with charge / discharge efficiency. According to Fig. 4, temperature is used as an explanatory variable, and charge / discharge efficiency is used as a response variable. (Estimation stage) Enter the temperature at the target time and estimate the charging / discharging efficiency.
[0047] <Second pattern> The storage battery is assumed to be connected to a solar power generation device. (Training stage) Training is performed using teacher data that associates the amount of solar power generation with the charge / discharge efficiency in the past. According to Fig. 4, the amount of solar power generation is set as an explanatory variable, and the charge / discharge efficiency is set as a response variable. (Estimation stage) The amount of solar power generation at the target time is input and the charging / discharging efficiency is estimated.
[0048] <Third pattern> The storage battery is assumed to be connected to a solar power generation device. (Training stage) Training is performed using teacher data that associates past temperature and solar power generation with charge / discharge efficiency. According to Fig. 4, temperature and solar power generation are used as explanatory variables, and charge / discharge efficiency is used as a response variable. (Estimation stage) The temperature and solar power generation amount at the target time are input to estimate the charging / discharging efficiency.
[0049] The "charge / discharge efficiency" specifically means "the ratio based on the amount of charge / discharge power and the amount of change in remaining power" in a storage battery. If the charging and discharging efficiency is high, the amount of electricity that increases (charges) in the storage battery is roughly the same as the amount of electricity received in the energy transaction. Also, the amount of electricity that decreases (discharges) in the storage battery is roughly the same as the amount of electricity to be transmitted in the energy transaction. On the other hand, if the charging / discharging efficiency is low, the amount of power that is added (charged) to the storage battery is less than the amount of power received through the energy transaction, and the amount of power that must be removed (discharged) from the storage battery is greater than the amount of power that should be transmitted through the energy transaction.
[0050] In addition, the relationship between the charge / discharge efficiency and the temperature and the amount of solar power generation is generally as follows: The higher the temperature and / or the more solar power generated, the higher the charging and discharging efficiency. The lower the temperature and / or the less solar power generated, the lower the charging and discharging efficiency. In this way, the charge / discharge efficiency can be estimated based on the air temperature around the indoor equipment 2 and the amount of power generated by the solar power generation device 26, without measuring the temperature of the storage battery itself.
[0051] <Charge / discharge efficiency -> Charge efficiency / Discharge efficiency> In another embodiment, the charge / discharge efficiency may be managed separately as a "charge efficiency" and a "discharge efficiency." That is, the "ratio of the amount of change in remaining power to the amount of charge power" in the storage battery is defined as the "charge efficiency," and the "ratio of the amount of discharge power to the amount of change in remaining power" is defined as the "discharge efficiency." In this case, the machine learning engine 10 is also controlled as follows. (Training stage) Training is performed separately using teacher data in which the charging efficiency and discharging efficiency are associated with the objective variable. (Estimation stage) The temperature and solar power generation amount at the target time are input, and the charging efficiency and discharging efficiency are estimated separately.
[0052] The remaining energy, the charging energy, and the discharging energy are transmitted as power usage information from the home gateway 23 to the estimation device 1. The remaining energy change is the amount of change in the remaining energy at the time when the charging energy or discharging energy is measured. For example, when the charging energy is from 12:00 to 12:30, the remaining energy change when calculating the charging efficiency is the difference between the remaining energy at 12:30 and the remaining energy at 12:00.
[0053] FIG. 5 is an explanatory diagram showing the determination of the amount of power in consideration of the charge / discharge efficiency when purchasing and selling electricity.
[0054] According to Fig. 5(a), when the remaining power of the storage battery is 6.0 kWh, the user purchases 2.0 kWh of electricity. If the remaining power of the storage battery is then 7.5 kWh and only 1.5 kWh has actually been charged, the charging efficiency is 75%. Charging capacity of purchased electricity: 2.0kWh Change in remaining battery power: 6.0kWh -> 7.5kWh Battery remaining power change: 1.5kWh (=7.5kWh-6.0kWh) Battery remaining power change = charge amount x charging efficiency 1.5KWh=2.0kWh×75% (75%=1.5kWh / 2.0kWh)
[0055] According to Fig. 5(b), when the remaining power of the storage battery was 7.5 kWh, the user sold 2.0 kWh of electricity. If the remaining power of the storage battery later becomes 5.5 kWh and only 1.0 kWh was actually discharged, the discharge efficiency was 50%. Discharge amount of electricity sold: 1.0kWh Change in remaining battery power: 7.5kWh -> 5.5kWh Battery remaining power change: 2.0kWh (=7.5kWh-5.5kWh) Battery remaining power change = discharge amount / charging efficiency 2.0KWh=1.0kWh / 50% (50%=1.0kWh / 2.0kWh) In this way, a difference occurs between the amount of change in the remaining power of the storage battery and the amount of charge and discharge power, and this difference is expressed as the charge and discharge efficiency. The HEMS gateway 222 (or the user) needs to determine the amount of power trading for the storage battery, taking into account the charge and discharge efficiency.
[0056] FIG. 6 is an explanatory diagram showing extended data in the training data of a machine learning engine.
[0057] According to the training data in Fig. 4, "temperature" and / or "amount of solar power generation" are used as explanatory variables. In contrast, according to the training data in Fig. 6, "time zone" classification and / or "weather" information are further included as explanatory variables. The "time period" may be 48 frames obtained by dividing a day into 30-minute intervals. Generally, the amount of solar radiation is highest around noon on a clear day, but the temperature tends to be highest around 2 p.m. It is also preferable to associate training data with each time period of a past day. As for the "weather" information, it is assumed that the temperature and the amount of solar power generation will differ between sunny and rainy days. The weather information can be obtained from the weather server 4, for example.
[0058] Moreover, according to the teacher data in FIG. 6, the explanatory variables further include the amount of charging power / the amount of discharging power during "control" or the amount of charging power / the amount of discharging power during "normal" time. The "controlled" state refers to an operating mode in which the storage battery transmits and receives (charges and discharges) power to the power grid. In addition, the normal time is, for example, a "green mode" and refers to an operation mode specific to the manufacturer, such as a clean mode, an energy saving mode, or an environmentally-friendly mode. For example, the green mode may be any mode other than the "control time". In the green mode, the surplus power generated by the photovoltaic power generation device after consumption within the home is charged to a storage battery, and the power charged by the storage battery is consumed at night. In this way, it is an operation mode that realizes local production and consumption of power as much as possible, and is effective for power resilience in the event of a large-scale power outage due to a disaster.
[0059] In this case, the machine learning engine is controlled as follows: (Training stage) Training is performed separately for charge / discharge efficiency during control and charge / discharge efficiency during normal operation. (Estimation step) The charge / discharge efficiency during control and the charge / discharge efficiency during normal operation are estimated separately. Then, the relationships between the amount of change in remaining power of the storage battery, the amount of charge and discharge, and the charge and discharge efficiency are calculated separately during control and during normal operation as follows. (When controlled) Battery remaining power change = charge amount x charging efficiency Battery remaining power change = discharge amount / discharge efficiency (normal time) Battery remaining power change = charge amount x charging efficiency Battery remaining power change = discharge amount / discharge efficiency
[0060] The charge / discharge power amount serving as the teacher data may be, for example, "the charge / discharge power amount during a predetermined period in the past (for example, the most recent few hours)." The charge / discharge power amount and the charge / discharge efficiency generally have the following relationship. The higher the charge / discharge power amount, the higher the charge / discharge efficiency. The lower the charge / discharge power amount, the lower the charge / discharge efficiency. The temperature of the storage battery affects the charge / discharge efficiency. The temperature of the storage battery is affected by the external environment (air temperature) and the charging / discharging operation of the storage battery itself (heat generation). Therefore, the amount of charge / discharge power from the last few hours affects the temperature of the storage battery itself, and also affects the charge / discharge efficiency.
[0061] Furthermore, the teacher data may include, for example, "the average charge / discharge efficiency for a specified period in the past (for example, the most recent past one month)." This represents the deterioration of the storage battery over time, and also affects the charge / discharge efficiency. Note that the deterioration over time differs depending on, for example, the storage battery ID and the storage battery capacity (rated capacity, rated input, rated output). The deterioration over time may be represented as the year the storage battery was installed.
[0062] <Control method based on maximum charge amount or maximum discharge amount during charging / discharging> Along with the charge / discharge efficiency, the maximum charge amount or maximum discharge amount during charging / discharging is estimated as follows. Maximum charge amount = min(Rated charge power amount, (Battery capacity - Remaining battery power amount) / Charge efficiency) Maximum discharge amount = min(Rated discharge power amount, Remaining battery power amount × Discharge efficiency) And the charge / discharge control device 211 of the in-house equipment device 2 is controlled based on the estimated maximum charge amount or maximum discharge amount. The maximum charge amount or maximum discharge amount may be used, for example, as the upper limit value of the bid power amount at the user terminal 24.
[0063] <Obtaining charge / discharge efficiency at the HEMS gateway> According to FIG. 2 described above, the HEMS gateway 222 connected to the battery 20 and the charge / discharge control device 211 communicates with the estimation device 1 via a network. (S1) The HEMS gateway 222 transmits the temperature and / or the solar power generation amount as environmental information to the estimation device 1. The environmental information is received by the environmental information receiving unit 11 of the estimation device 1 described above, and the environmental information is input to the machine learning engine 10. (S2) On the other hand, the estimation device 1 estimates the "charge / discharge efficiency" by the machine learning engine 10 according to the environmental information, and returns the charge / discharge efficiency to the HEMS gateway 222. The charge / discharge efficiency is transmitted to the HEMS gateway 222 of the in-house equipment device 2 by the charge / discharge efficiency transmitting unit 12 of the estimation device 1 described above. (S3) As a result, the HEMS gateway 222 can control the charge / discharge of the battery 20 in consideration of the received charge / discharge efficiency. For example, as shown in FIG. 5 described above, the charge / discharge power amount can be determined in consideration of the charge / discharge efficiency.
[0064] <Estimation of bid power amount in power trading> According to FIG. 5 described above, it can also be applied at the time of power purchase and power sale. For example, it is useful when determining the charge power amount or discharge power amount to bid during the power trading bid. In the case of a storage battery 20 with low charge / discharge efficiency, it is necessary to estimate the amount of charging energy to be bid on higher than necessary. Otherwise, it may happen that the amount of energy charged to the storage battery is less than the amount of charging energy bid. On the other hand, in the case of a storage battery 20 with low charge / discharge efficiency, it is necessary to estimate the amount of discharged power to be bid on low. Otherwise, it may happen that the amount of power discharged to the power grid is less than the amount of power discharged from the storage battery.
[0065] As described above in detail, the present invention Systems and methods According to the method, the charge / discharge efficiency of the storage battery can be estimated. This makes it possible to estimate the charge / discharge efficiency according to the ambient temperature of the home equipment and the amount of power generated by the photovoltaic power generation device, without measuring the temperature of the storage battery itself.
[0066] Furthermore, this will make it possible, for example, to "improve the accuracy of electricity volumes in electricity trading by estimating the charging and discharging efficiency of storage batteries," thereby contributing to Goal 7 of the United Nations-led Sustainable Development Goals (SDGs) to "ensure access to affordable, reliable, sustainable and modern energy."
[0067] Regarding the various embodiments of the present invention described above, various changes, modifications, and omissions within the scope of the technical idea and perspective of the present invention can be easily made by those skilled in the art. The above description is merely an example and is not intended to be restrictive in any way. The present invention is limited only by the scope of the claims and their equivalents. [Explanation of symbols]
[0068] 1 Estimation device 10 Machine Learning Engine 11 Environmental Information Reception Department 12 Charging and discharging efficiency transmitter 2. Home equipment 20. Storage Battery 211 Charge / discharge control device 212 Distribution Board 221 Smart Meter 222 HEMS Gateway 23 Home Gateway 24 User terminals 25 Thermometer 26 Photovoltaic power generation equipment 3 Power transmission and distribution equipment 4. Weather Server
Claims
1. A system in which an estimation device that estimates a charge / discharge efficiency of a storage battery and a home equipment device having a storage battery and a charge / discharge control device communicate with each other via a network, The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is trained in a training stage using teacher data in which past temperatures are associated with charge / discharge efficiency, and in an estimation stage, the estimation device has a machine learning engine that inputs the temperature at a target time and estimates the charge / discharge efficiency, and transmits the estimated charge / discharge efficiency to the home equipment device. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. A system characterized by:
2. A system in which an estimation device that estimates a charge / discharge efficiency of a storage battery connected to a photovoltaic power generation device and a home equipment device having a storage battery and a charge / discharge control device communicate with each other via a network, The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is trained in a training stage using teacher data in which past photovoltaic power generation amounts and charge / discharge efficiency are associated with each other, and in an estimation stage, the estimation device has a machine learning engine that inputs the photovoltaic power generation amount at a target time and estimates the charge / discharge efficiency, and transmits the estimated charge / discharge efficiency to the home equipment. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. A system characterized by:
3. A system in which an estimation device that estimates a charge / discharge efficiency of a storage battery connected to a photovoltaic power generation device and a home equipment device having a storage battery and a charge / discharge control device communicate with each other via a network, The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is trained in a training stage using teacher data in which past temperatures and amounts of photovoltaic power generation are associated with charge / discharge efficiency, and in an estimation stage, the estimation device has a machine learning engine that inputs the temperature and amount of photovoltaic power generation at a target time and estimates the charge / discharge efficiency, and transmits the estimated charge / discharge efficiency to the home equipment. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. A system characterized by:
4. Regarding the charge / discharge efficiency, the ratio of the amount of change in the remaining power to the amount of charge power is defined as the charge efficiency, and the ratio of the amount of discharge power to the amount of change in the remaining power is defined as the discharge efficiency. The machine learning engine of the estimation device separately trains the charging efficiency and the discharging efficiency using teacher data in which the charging efficiency and the discharging efficiency are associated with the objective variable in a training stage, and separately estimates the charging efficiency and the discharging efficiency in an estimation stage.
4. The system according to claim 1, wherein the first and second inputs are connected to a first input port.
5. The teacher data further includes time zones and / or weather information as explanatory variables.
5. A system according to claim 1, wherein the first and second inputs are connected to a first input port.
6. The teacher data further includes, as explanatory variables, the charge power amount / discharge power amount during control or the charge power amount / discharge power amount during normal operation, The machine learning engine of the estimation device is In the training stage, training is performed separately for the charge / discharge efficiency during control and the charge / discharge efficiency during normal operation. In the estimation step, the charge / discharge efficiency during control and the charge / discharge efficiency during normal operation are estimated separately.
6. A system according to any one of claims 1 to 5.
7. Along with the charge / discharge efficiency, the maximum charge amount or maximum discharge amount during charging / discharging is estimated as follows: Maximum charging capacity = min (rated charging capacity, (battery capacity - remaining battery capacity) / charging efficiency) Maximum discharge amount = min (rated discharge amount, remaining battery power amount x discharge efficiency) 7. A system according to any one of claims 1 to 6.
8. An estimation method for a system in which an estimation device that estimates a charge / discharge efficiency of a storage battery and a home equipment device having a storage battery and a charge / discharge control device communicate with each other via a network, comprising: The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, In a training stage, the estimation device is trained using teacher data in which past temperatures are associated with charge / discharge efficiency, and in an estimation stage, the temperature at a target time is input, and the estimated charge / discharge efficiency is transmitted to the home equipment using a machine learning engine that estimates the charge / discharge efficiency. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. The estimation method according to claim 1,
9. An estimation method for a system in which an estimation device that estimates a charge / discharge efficiency of a storage battery connected to a photovoltaic power generation device and a home equipment device having a storage battery and a charge / discharge control device communicate with each other via a network, comprising: The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, In a training stage, the estimation device is trained using teacher data in which past photovoltaic power generation amounts and charge / discharge efficiency are associated with each other, and in an estimation stage, the estimation device inputs the photovoltaic power generation amount at a target time, and uses a machine learning engine that estimates the charge / discharge efficiency to transmit the estimated charge / discharge efficiency to the home equipment; The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. The estimation method according to claim 1,
10. An estimation method for a system in which an estimation device that estimates a charge / discharge efficiency of a storage battery connected to a photovoltaic power generation device and a home equipment device having a storage battery and a charge / discharge control device communicate with each other via a network, comprising: The charge / discharge efficiency is determined as a ratio based on the charge / discharge power amount and the remaining power change amount in the storage battery, The estimation device is trained in a training stage using teacher data in which past temperatures and amounts of photovoltaic power generation are associated with charge / discharge efficiency, and in an estimation stage, the estimation device has a machine learning engine that inputs the temperature and amount of photovoltaic power generation at a target time and estimates the charge / discharge efficiency, and transmits the estimated charge / discharge efficiency to the home equipment. The charge / discharge control device of the home equipment controls charging / discharging of the storage battery based on the received charge / discharge efficiency. The estimation method according to claim 1,
Citation Information
Patent Citations
Controller for DC power supply, DC power supply system, and control method for controller for DC power supply
JP2016100956A
Power management device
JP2017022918A
Control device, power control system, control method, and program
JP2020156149A
Program, device, and method for improving the accuracy of predicting future power consumption using location information of mobile terminals
JP6714489B2
Program, device, and method for predicting future power consumption using an unspecified number of posts
JP6812320B2
