Intelligent control method and system for power distribution cabinet, power distribution cabinet and storage medium

Abstract

The invention provides an intelligent control method and system for a power distribution cabinet, the power distribution cabinet and a storage medium, and belongs to the technical field of power distribution, the method is applied to an energy storage system, the energy storage system comprises the power distribution cabinet and an energy storage element, the energy storage element is connected with a power grid in parallel through the power distribution cabinet, and the method is executed by the power distribution cabinet. The method comprises the steps of determining a load prediction value of a power grid in a future time period based on historical load data of the power grid; determining the target charging and discharging power of the energy storage element based on the load prediction value of the power grid in the future time period and the current residual electric quantity of the energy storage element; and controlling the energy storage element to charge or discharge a power grid according to the target charge and discharge power based on the target charge and discharge power. According to the power distribution cabinet intelligent control method and system, the power distribution cabinet and the storage medium, the stability of power grid operation can be improved.

Description

Technical Field

[0001] This application belongs to the technical field of power distribution, and more specifically, relates to an intelligent control method and system for a power distribution cabinet, a power distribution cabinet, and a storage medium. Background Art

[0002] An energy storage system can store electrical energy and release it when needed to regulate energy supply and demand, improve the stability and efficiency of the power system, and its functions widely cover multiple fields such as the power system, renewable energy, and the user side. As a key power control device in the energy storage system, the power distribution cabinet is mainly used for the distribution, conversion, protection, and monitoring of electrical energy.

[0003] Existing power distribution cabinet control methods mostly adopt a fixed threshold control method, where charging stops when the energy storage device's power reaches the upper limit and discharging stops when it reaches the lower limit. The fixed threshold cannot adapt to the dynamic changes of the power grid, which may lead to poor coordination between the energy storage system and the power grid and affect the stability of the entire power grid operation. Summary of the Invention

[0004] The purpose of this application is to provide an intelligent control method and system for a power distribution cabinet, a power distribution cabinet, and a storage medium to improve the stability of power grid operation.

[0005] In the first aspect of the embodiments of this application, an intelligent control method for a power distribution cabinet is provided, which is applied to an energy storage system. The energy storage system includes a power distribution cabinet and energy storage elements. The energy storage elements are connected in parallel with the power grid through the power distribution cabinet. The method is executed by the power distribution cabinet, and the method includes: Determining a load prediction value of the power grid in a future period based on historical load data of the power grid; the load prediction value of the power grid in the future period includes N load prediction values of the power grid in the next N periods; Determining the target charge-discharge power of the energy storage element based on the load prediction value of the power grid in the future period and the current remaining power of the energy storage element; Controlling the energy storage element to charge or discharge the power grid according to the target charge-discharge power based on the target charge-discharge power; Among them, the determining the target charge-discharge power of the energy storage element based on the load prediction value of the power grid in the future period and the current remaining power of the energy storage element includes: Inputting the load prediction value of the power grid in the future period and the current remaining power of the energy storage element into a reinforcement learning network to obtain the target charge-discharge power of the energy storage element; Among them, the action space of the reinforcement learning network is constructed based on the theoretical value range of the charge-discharge power, and each action in the action space includes a charge-discharge power candidate value of the energy storage element in each of the next N periods.

[0006] In the second aspect of the embodiments of the present application, an intelligent control system for a power distribution cabinet is provided, which is arranged in the power distribution cabinet. The power distribution cabinet is included in an energy storage system, and the energy storage system further includes energy storage elements. The energy storage elements are connected in parallel with the power grid through the power distribution cabinet. The intelligent control system for the power distribution cabinet includes: A grid load prediction module, configured to determine the load prediction value of the power grid in a future time period based on the historical load data of the power grid; the load prediction value of the power grid in the future time period includes N load prediction values of the power grid in the next N time periods; A charge-discharge power determination module, configured to determine the target charge-discharge power of the energy storage element based on the load prediction value of the power grid in the future time period and the current remaining power of the energy storage element; A control module, configured to control the energy storage element to charge or discharge the power grid according to the target charge-discharge power based on the target charge-discharge power; Specifically, the charge-discharge power determination module is configured to: Input the load prediction value of the power grid in the future time period and the current remaining power of the energy storage element into a reinforcement learning network to obtain the target charge-discharge power of the energy storage element; Wherein, the action space of the reinforcement learning network is constructed based on the theoretical value range of the charge-discharge power, and each action in the action space includes the charge-discharge power candidate value of the energy storage element in each of the next N time periods.

[0007] In the third aspect of the embodiments of the present application, a power distribution cabinet is provided, which includes a controller. The controller includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the above-mentioned intelligent control method for the power distribution cabinet are implemented.

[0008] In the fourth aspect of the embodiments of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned intelligent control method for the power distribution cabinet are implemented.

[0009] The beneficial effects of the intelligent control method and system for the power distribution cabinet, the power distribution cabinet, and the storage medium provided by the embodiments of the present application are as follows: Based on the predicted change of the power grid load and the current remaining power of the energy storage element, the embodiments of the present application determine the target charge-discharge power (i.e., the optimal charge-discharge power) of the energy storage element in the current time period, and control the energy storage element to charge or discharge the power grid based on the target charge-discharge power to adapt to the change of the power grid load, which can enhance the coordination between the energy storage system and the power grid and ensure the stable operation of the power grid. Description of the Drawings

[0010] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0011] Figure 1 The structural block diagram of an energy storage system provided by an embodiment of the present application; Figure 2 The schematic flowchart of the intelligent control method for a power distribution cabinet provided by an embodiment of the present application; Figure 3 The structural block diagram of the intelligent control system for a power distribution cabinet provided by an embodiment of the present application; Figure 4 The schematic block diagram of a controller provided by an embodiment of the present application. Detailed implementation manners

[0012] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0013] To make the objectives, technical solutions, and advantages of the present application clearer, the following will be described through specific embodiments with reference to the drawings.

[0014] An embodiment of the present application provides an intelligent control method for a power distribution cabinet, which is applied to an energy storage system 110. As Figure 1 shown, the energy storage system 110 includes a power distribution cabinet 112 and energy storage elements 111. The energy storage elements 111 are connected in parallel with the power grid 120 through the power distribution cabinet 112.

[0015] Further, an intelligent control method for a power distribution cabinet provided by an embodiment of the present application can be executed by Figure 1 the power distribution cabinet 112 shown in Figure 2 , Figure 2 The schematic flowchart of the intelligent control method for a power distribution cabinet provided by an embodiment of the present application. The method may include: S101: Determine the load prediction value of the power grid in a future period based on the historical load data of the power grid; the load prediction value of the power grid in the future period includes N load prediction values of the power grid in the next N periods; In this embodiment, the historical load data of the power grid can reflect the electricity consumption habits of users (such as the day-night fluctuations of industrial loads and the morning-evening peaks of residential loads). Therefore, the load of the power grid in a future period can be predicted based on the historical load data of the power grid, and the load prediction value of the power grid in the future period can be obtained. Among them, the future period can be several hours to several days in the future. For example, the load of the power grid in the next 1 hour, 2 hours, or 1 day, 2 days can be predicted.

[0016] Specifically, since the power grid supplies power in regions, by separately setting energy storage systems in one or more power supply regions, the stable operation of the power grid in the corresponding power supply regions can be achieved. Therefore, in this embodiment, the historical load data of the power grid refers to the historical load data of the power grid in any power supply region (denoted as the target power supply region). The historical load data of the power grid in the target power supply region can be obtained from the power grid dispatching center responsible for this power supply region. For example, the historical load data of the power grid in the most recent 1 month can be obtained from the power grid dispatching center of the target power supply region, and the historical load data of the power grid is input into a pre-trained eXtreme Gradient Boosting (XGBoost) model or a Light Gradient Boosting Machine (LightGBM) model to obtain the load prediction value of the power grid in the future period (7 days).

[0017] S102: Determine the target charge-discharge power of the energy storage element based on the load prediction value of the power grid in the future period and the current remaining power of the energy storage element.

[0018] In this embodiment, the energy storage element can be a battery or a supercapacitor, etc. Based on obtaining the load prediction value of the power grid in the future period, if the load prediction value of the power grid in the future period is greater than the current power grid load, the energy storage element can be controlled to discharge to provide electric energy for the power grid and relieve the power supply pressure of the power grid; if the load prediction value of the power grid in the future period is less than the current power grid load, the energy storage element can be charged to absorb excess electric energy and avoid power grid voltage / frequency fluctuations.

[0019] For example, if it is predicted that the power grid load will increase within the next 1 hour (such as during the evening peak), the energy storage element can be controlled to discharge in advance during the current period to relieve the upcoming power supply pressure. Another example is that if it is predicted that there will be a load trough within the next 2 hours (such as late at night), the energy storage element can be controlled to charge during the current period to absorb excess electric energy.

[0020] Specifically, considering the limitation of the current remaining power of the energy storage element, to avoid overcharging or over-discharging of the energy storage element, the target charge-discharge power of the energy storage element in the current period can be determined by combining the load prediction value of the power grid in the future period and the current remaining power of the energy storage element, and the energy storage element is controlled to charge or discharge at the target charge-discharge power. In this way, damage to the energy storage element can be reduced on the premise of meeting the power grid demand. Among them, the current remaining power of the energy storage element can be obtained by reading the data of the corresponding management system. For example, if the energy storage element is a battery, the current remaining power of the battery can be obtained by reading the data of the battery management system; if the energy storage element is a supercapacitor, the current remaining power of the supercapacitor can be obtained by reading the data of the supercapacitor management system.

[0021] S103: Control the energy storage element to charge or discharge the power grid at the target charge-discharge power based on the target charge-discharge power.

[0022] In this embodiment, a controller and a power conversion system (PCS) are usually arranged in the power distribution cabinet. The power conversion system can realize the conversion of electrical energy between the alternating current of the power grid and the direct current of the energy storage element. The controller adjusts the output of the power conversion system according to the target charge-discharge power, and can adjust the charge-discharge current / voltage of the energy storage element to realize the tracking control of the target charge-discharge power.

[0023] It can be concluded from the above that this embodiment determines the target charge-discharge power (i.e., the optimal charge-discharge power) of the energy storage element in the current period based on the predicted change of the power grid load and the current remaining power of the energy storage element, and controls the energy storage element to charge or discharge the power grid based on the target charge-discharge power to adapt to the change of the power grid load, which can enhance the coordination between the energy storage system and the power grid and ensure the stable operation of the power grid.

[0024] In an embodiment of the present application, based on the load prediction value of the power grid in the future period and the current remaining power of the energy storage element, determining the target charge-discharge power of the energy storage element may specifically include: Input the load prediction value of the power grid in the future period and the current remaining power of the energy storage element into the reinforcement learning network to obtain the target charge-discharge power of the energy storage element; Among them, the action space of the reinforcement learning network is constructed based on the theoretical value range of the charge-discharge power, and each action in the action space includes the candidate charge-discharge power value of the energy storage element in each of the next N periods.

[0025] Specifically, the calculation process of the reward function of the reinforcement learning network includes: Determine the power grid load fluctuation value based on the N load prediction values of the power grid in the next N periods; Based on the candidate charge and discharge power values of the energy storage element in each of the next N time periods and the current remaining power of the energy storage element, determine N predicted remaining power values of the energy storage element in the next N time periods; Based on the N predicted remaining power values of the energy storage element in the next N time periods, determine the state of charge evaluation value of the energy storage element; Based on the grid load fluctuation value and the state of charge evaluation value of the energy storage element, calculate the reward function of the reinforcement learning network.

[0026] In this embodiment, by predicting the load values of the power grid in the next N time periods, the charge and discharge strategy of the energy storage element can be determined in advance, globally optimizing the multi-time period objective and avoiding short-sighted decisions in a single time period.

[0027] Specifically, a reinforcement learning network can be used to determine the target charge and discharge power of the energy storage element. The reinforcement learning network optimizes the decision-making strategy through the dynamic interaction between the agent and the environment (the state space determined by the input data), tries new actions, uses the reward function feedback to evaluate the action value, and gradually learns the optimal mapping from the state to the action.

[0028] Among them, each action corresponds to a charge and discharge power sequence for the next N time periods , ,..., ..., , and each power value in this sequence is respectively the candidate charge and discharge power value of the energy storage element in each of the next N time periods. For example, is the candidate charge and discharge power value for the i-th time period in the future (a positive number represents discharge, and a negative number represents charge). The subscript t represents the current time period. At the same time, the theoretical value range of the charge and discharge power determines the maximum and minimum values of the allowable charge and discharge power of the energy storage element, and each power value in the charge and discharge power sequence should be selected within the theoretical value range. By calculating the reward function corresponding to each action, select the target action from the action space, and determine the target charge and discharge power based on the charge and discharge power sequence corresponding to the target action.

[0029] Exemplarily, the reward function corresponding to each action is determined according to the grid load fluctuation value and the state of charge evaluation value of the energy storage element. The larger the grid load fluctuation value, the greater the degree of grid load fluctuation, and the smaller the value of the reward function; the state of charge evaluation value of the energy storage element is used to characterize the safety degree of the charge and discharge process of the energy storage element. The larger the state of charge evaluation value of the energy storage element, the lower the safety degree of the charge and discharge process of the energy storage element, and the smaller the value of the reward function.

[0030] Specifically, the power grid load fluctuation value can be determined based on the N load prediction values of the power grid in the next N time periods. Correspondingly, based on the charge and discharge power candidate values of the energy storage element in each of the next N time periods and the current remaining power of the energy storage element, the N remaining power prediction values of the energy storage element in the next N time periods can be determined. The calculation formula for the remaining power prediction value of the energy storage element in the i-th time period in the future is as follows: ; Wherein, represents the predicted remaining power value of the energy storage element in the i-th time period in the future, represents the predicted remaining power value of the energy storage element in the (i - 1)-th time period in the future, is the charge and discharge efficiency of the energy storage element, represents the candidate charge and discharge power value in the (i - 1)-th time period in the future, represents the rated capacity of the energy storage element, represents the time difference between the i-th time period and the (i - 1)-th time period in the future.

[0031] Based on the N remaining power prediction values of the energy storage element in the next N time periods, the charge and discharge performance of the energy storage element can be evaluated to obtain the state of charge evaluation value of the energy storage element.

[0032] Furthermore, based on the power grid load fluctuation value and the state of charge evaluation value of the energy storage element, the two can be weighted and summed, and the reward function of the reinforcement learning network can be determined based on the result of the weighted sum.

[0033] It can be concluded from the above that this embodiment is based on the N load prediction values of the power grid in the next N time periods, and uses a reinforcement learning network for multi-time period joint optimization, which is beneficial to determining the optimal target charge and discharge power.

[0034] In an embodiment of the present application, determining the power grid load fluctuation value based on the N load prediction values of the power grid in the next N time periods may specifically include: Obtain the target load curve of the power grid in the next N time periods; Based on the N load prediction values of the power grid in the next N time periods and the charge and discharge power candidate values of the energy storage element in each of the next N time periods, determine the N net load prediction values of the power grid in the next N time periods; wherein, the net load prediction value is the load prediction value of the power grid itself; Compare the N net load prediction values of the power grid in the next N time periods with the target load curve of the power grid in the next N time periods to obtain the power grid load fluctuation value.

[0035] In this embodiment, those skilled in the art can preset the target load curve of the power grid for a future period according to factors such as historical load data, industry electricity consumption patterns, meteorological forecasts (such as the impact of temperature on air-conditioning load), and holidays. During the operation of the power grid, the power source (such as a generator) is usually controlled in advance according to the target load curve of the power grid to adjust the output, avoiding power outages or surpluses. The curve corresponding to the corresponding period can be intercepted from the target load curve of the power grid to obtain the target load curve of the power grid for the next N periods.

[0036] The predicted value of the net load of the power grid refers to the actual load value of the power grid after considering the impact of energy storage charging and discharging. That is, if the candidate value of the charging and discharging power of the energy storage element is greater than zero in the i-th period in the future, it indicates that the energy storage element discharges power to the power grid during this period. At this time, the predicted value of the net load of the power grid is equal to the predicted load value of the power grid minus the discharging power of the energy storage element; if the candidate value of the charging and discharging power of the energy storage element is less than zero in the i-th period in the future, it indicates that the power grid charges the energy storage element during this period. At this time, the predicted value of the net load of the power grid is equal to the predicted load value of the power grid plus the discharging power of the energy storage element.

[0037] In this embodiment, comparing the N predicted values of the net load of the power grid for the next N periods with the target load curve of the power grid for the next N periods can be described in detail as follows: Determine the target load value corresponding to each period in the next N periods based on the load value of the corresponding period on the target load curve of the power grid for the next N periods; Compare the N predicted values of the net load of the power grid for the next N periods with the target load values of the corresponding periods to obtain the power grid load fluctuation value. The specific calculation formula is as follows: ; Where represents the power grid load fluctuation value, represents the predicted load value of the power grid in the i-th period in the future, represents the candidate value of the charging and discharging power of the energy storage element in the i-th period in the future, represents the predicted value of the net load of the power grid in the i-th period in the future, represents the target load value corresponding to the i-th period in the future on the target load curve.

[0038] It can be concluded from the above that in this embodiment, by calculating N net load prediction values of the power grid in the next N time periods, and the target load value corresponding to each time period in the next N time periods, and comparing the N net load prediction values of the power grid in the next N time periods with the target load values of the corresponding time periods, the power grid load fluctuation value is obtained, and the reward function of the reinforcement learning network is determined based on the power grid load fluctuation value, which is beneficial to guiding the reinforcement learning network to determine the target charge and discharge power with the goal of "reducing the difference between the N net load prediction values and the target load values of the corresponding time periods", so that the actual net load curve of the power grid is closer to the target load curve, which is beneficial to the stable operation of the power grid.

[0039] In one embodiment of the present application, determining the state of charge evaluation value of the energy storage element based on the N remaining charge prediction values of the energy storage element in the next N time periods includes: Obtaining the charge and discharge safety threshold of the energy storage element; Comparing the N remaining charge prediction values of the energy storage element in the next N time periods with the charge and discharge safety threshold to determine the charge and discharge risk evaluation value of the energy storage element in the next N time periods; Determining the change rate of the remaining charge between two adjacent time periods based on the N remaining charge prediction values of the energy storage element in the next N time periods, obtaining N - 1 remaining charge change rates, and determining the fluctuation value of the remaining charge based on the N - 1 remaining charge change rates; Performing weighted summation on the charge and discharge risk evaluation value of the energy storage element in the next N time periods and the fluctuation value of the remaining charge to obtain the state of charge evaluation value of the energy storage element.

[0040] In this embodiment, taking the battery as an example, the charge and discharge safety threshold of the energy storage element may include the upper limit of the remaining charge (such as 95%) and the lower limit of the remaining charge (such as 5%). If the battery exceeds the charge and discharge safety threshold during the charge and discharge process, it may lead to shortened life (such as lithium plating, sulfidation) or safety risks (such as thermal runaway).

[0041] Therefore, in this embodiment, by comparing the N remaining charge prediction values of the energy storage element in the next N time periods with the charge and discharge safety threshold, the charge and discharge risk evaluation value of the energy storage element in the next N time periods is obtained, and the state of charge evaluation value of the energy storage element is determined based on the charge and discharge risk evaluation value.

[0042] Specifically, the following formula can be used to calculate the charge and discharge risk evaluation value: ; Wherein, represents the charge and discharge risk evaluation value of the energy storage element in the next N time periods, represents the charge and discharge risk evaluation value of the i-th time period, represents the remaining charge prediction value of the i-th time period, Represents the upper threshold value for charge and discharge safety, Represents the lower threshold value for charge and discharge safety, and Represents a preset proportionality coefficient.

[0043] The above formula indicates that for any future time period i, if the predicted remaining power value corresponding to it is between and , then the charge and discharge risk assessment value for the i-th time period is equal to 0; if the predicted remaining power value is greater than , then based on the difference between and , the charge and discharge risk assessment value for the i-th time period is determined; if the predicted remaining power value is less than , then based on the difference between and , the charge and discharge risk assessment value for the i-th time period is determined. According to the above method, the charge and discharge risk assessment values for each time period can be obtained.

[0044] Based on the charge and discharge risk assessment values for each time period obtained, the charge and discharge risk assessment values for each time period can be accumulated to obtain the charge and discharge risk assessment value of the energy storage element in the future N time periods.

[0045] At the same time, considering that if the change rate of the remaining power between two adjacent time periods is too large, it may cause the battery to heat up more, the internal resistance to increase, and accelerate aging. Therefore, in this embodiment, the state of charge assessment value of the energy storage element is also determined based on the change rate of the remaining power.

[0046] Specifically, based on the change rate of the remaining power between two adjacent time periods in the future N time periods, N - 1 change rates of the remaining power can be obtained, and based on the N - 1 change rates of the remaining power, the fluctuation value of the remaining power can be determined. For example, the following formula can be used to calculate the fluctuation value of the remaining power: ; where, represents the fluctuation value of the remaining power, represents the predicted remaining power value for the i-th future time period, represents the predicted remaining power value for the (i + 1)-th future time period.

[0047] Based on the charge and discharge risk assessment value and the fluctuation value of the remaining power obtained, the charge and discharge risk assessment value and the fluctuation value of the remaining power can be weighted and summed to obtain the state of charge assessment value of the energy storage element.

[0048] As can be seen from the above, in this embodiment, by comprehensively considering the charge and discharge risk assessment value and the fluctuation value of the remaining power, the state of charge assessment value of the energy storage element is determined, which can optimize the operation efficiency of the energy storage element on the basis of ensuring that the energy storage element does not exceed the safety boundary.

[0049] In an embodiment of the present application, calculating a reward function of a reinforcement learning network based on a grid load fluctuation value and a state of charge assessment value of an energy storage element includes: Determining a first weight of the grid load fluctuation value based on the proportion of new energy power generation in the grid; wherein, when the proportion of renewable energy power generation in the grid is greater than a first threshold, the corresponding first weight is greater than the first weight corresponding to the proportion of renewable energy power generation in the grid being less than or equal to the first threshold; Determining a second weight of the state of charge assessment value of the energy storage element based on the first weight; Performing weighted summation on the grid load fluctuation value and the state of charge assessment value of the energy storage element based on the first weight and the second weight to obtain a first summation result; Determining the reciprocal of the first summation result as the reward function of the reinforcement learning network.

[0050] In this embodiment, considering that the larger the grid load fluctuation value and the state of charge assessment value of the energy storage element are, the smaller the corresponding reward function is. Therefore, on the basis of obtaining the grid load fluctuation value and the state of charge assessment value of the energy storage element, the two can be weighted and summed, and the reciprocal of the first summation result obtained by weighted summation (in order to avoid the denominator being zero, the reciprocal can be taken after adding 1 to the first summation result) is used as the reward function of the reinforcement learning network. Among them, the first weight of the grid load fluctuation value and the second weight of the state of charge assessment value form a complementary relationship, and their sum is 1.

[0051] Furthermore, considering that when the proportion of renewable energy (such as wind power and photovoltaic) power generation in the grid is relatively large, due to the strong intermittency of the output of renewable energy, the grid load volatility will be aggravated. At this time, the weight of the grid load fluctuation value can be increased to prompt the reinforcement learning network to give priority to suppressing fluctuations and ensure the stability of the grid. Among them, the proportion of renewable energy power generation in the grid can be obtained based on the real-time operation data of the grid dispatching center.

[0052] Specifically, a first threshold can be preset. When the proportion of the power generation of renewable energy (such as wind power and photovoltaic power) in the power grid is greater than the first threshold, it indicates that the proportion of the power generation of renewable energy in the power grid is relatively large. At this time, the first weight of the power grid load fluctuation value is set to a relatively large first value; when the proportion of the power generation of renewable energy (such as wind power and photovoltaic power) in the power grid is less than or equal to the first threshold, it indicates that the proportion of the power generation of renewable energy in the power grid is relatively small. At this time, the first weight of the power grid load fluctuation value is set to a relatively small second value to increase the second weight of the state of charge evaluation value and prompt the reinforcement learning network to give priority to ensuring the healthy operation of the energy storage element.

[0053] It can be concluded from the above that based on the proportion of the power generation of renewable energy in the power grid, this embodiment dynamically adjusts the first weight of the power grid load fluctuation value, enabling the energy storage system to achieve the best adjustment effect in different scenarios and enhancing the collaborative operation effect of the power grid and the energy storage system.

[0054] In an embodiment of the present application, inputting the load prediction value of the power grid in a future period and the current remaining power of the energy storage element into the reinforcement learning network to obtain the target charge and discharge power of the energy storage element includes: Inputting the load prediction value of the power grid in a future period and the current remaining power of the energy storage element into the reinforcement learning network to obtain the target action; Determining the charge and discharge power candidate value of the target period among the N charge and discharge power candidate values included in the target action as the target charge and discharge power of the energy storage element, where the target period is the period closest to the current period within the N periods.

[0055] In this embodiment, when the load prediction value of the power grid in a future period and the current remaining power of the energy storage element are input into the reinforcement learning network, among the target actions output by the reinforcement learning network, the charge and discharge power candidate value of each period is the optimal value of the charge and discharge power of each period. On this basis, select the period closest to the current period (for example, if the current is the t-th moment, the target period is t + 1), use this period as the target period, and use the charge and discharge power candidate value of the target period as the target charge and discharge power to control the charge and discharge of the energy storage element in the current period. The power candidate values of other periods (such as t + 2 to t + N) can be dynamically adjusted in combination with newly collected data and used as the target charge power for the next period.

[0056] It can be concluded from the above that this embodiment calculates the reward function based on the prediction information of the future N periods and determines the target charge and discharge power of the current period, which can ensure the real-time nature of the charge and discharge control of the energy storage element on the basis of achieving the long-term optimization goal.

[0057] In an embodiment of the present application, the method for determining the theoretical value range of the charge and discharge power includes: Obtaining the initial value range; Determine a first adjustment coefficient based on the temperature of the energy storage element; Determine a second adjustment coefficient based on the remaining life of the energy storage element; Based on the minimum adjustment coefficient among the first adjustment coefficient and the second adjustment coefficient, adjust the initial value range to obtain the theoretical value range of the charge-discharge power.

[0058] In this embodiment, the initial value range determines the initial safety boundary of the charge-discharge power of the energy storage element, and the initial value range can be obtained according to the design parameters of the energy storage element.

[0059] Considering that the temperature of the energy storage element will affect its charge-discharge performance, taking the battery as an example, overcharge and over-discharge in a high-temperature environment may lead to thermal runaway and accelerate battery aging, and large-current charge and discharge in a low-temperature environment will reduce the capacity utilization rate and increase the internal resistance loss.

[0060] At the same time, considering that the remaining life of the energy storage element can reflect its state of health (SOH), the internal resistance of the battery near the end of its life increases, the capacity decays, and the ability to withstand large-current charge and discharge decreases, which is prone to cause failures.

[0061] Therefore, in this embodiment, based on the initial value range, a first adjustment coefficient is determined based on the temperature of the energy storage element, a second adjustment coefficient is determined based on the remaining life of the energy storage element, and then the initial value range is adjusted based on the minimum adjustment coefficient among the first adjustment coefficient and the second adjustment coefficient to obtain the theoretical value range of the charge-discharge power.

[0062] For example, the initial value range is [-50kW, 50kW] (negative value for charging and positive value for discharging). If the first adjustment coefficient determined according to the temperature of the energy storage element is 0.9 and the second adjustment coefficient determined according to the remaining life of the energy storage element is 0.7, then the initial value range is adjusted based on the second adjustment coefficient to obtain the theoretical value range of the charge-discharge power as [-35kW, 35kW].

[0063] Specifically, the first adjustment coefficient can be determined by the following first formula, and the first formula is: ; Wherein, represents the first adjustment coefficient, represents the current temperature of the energy storage element, represents the reference value of the temperature, represents the lower limit value of the temperature, represents the upper limit value of the temperature, and both represent preset proportional coefficients.

[0064] At the same time, the second adjustment coefficient can be determined by the following second formula, and the second formula is: ; Wherein, represents the second adjustment coefficient, represents the remaining life of the energy storage element, represents the reference value of the remaining life, represents a preset proportionality coefficient.

[0065] It can be obtained from the above that this embodiment dynamically adjusts the theoretical value range of the charge-discharge power based on the actual performance of the energy storage element, further improving the collaborative operation effect of the power grid and the energy storage system.

[0066] Corresponding to the intelligent control method of the distribution cabinet in the above embodiment, Figure 3 is the structural block diagram of the intelligent control system of the distribution cabinet provided by an embodiment of the present application. For the convenience of description, only the parts related to the embodiment of the present application are shown. Refer to Figure 3 , the intelligent control system 20 of the distribution cabinet is set in the distribution cabinet 112. The distribution cabinet 112 is included in the energy storage system 110. The energy storage system 110 further includes an energy storage element 111. The energy storage element 111 is connected in parallel with the power grid 120 through the distribution cabinet 112. The intelligent control system 20 of the distribution cabinet includes: a power grid load prediction module 21, a charge-discharge power determination module 22, and a control module 23. Wherein, the power grid load prediction module 21 is used to determine the load prediction value of the power grid in the future time period based on the historical load data of the power grid; the load prediction value of the power grid in the future time period includes N load prediction values of the power grid in the next N time periods; The charge-discharge power determination module 22 is used to determine the target charge-discharge power of the energy storage element based on the load prediction value of the power grid in the future time period and the current remaining power of the energy storage element; The control module 23 is used to control the energy storage element to charge or discharge the power grid according to the target charge-discharge power based on the target charge-discharge power.

[0067] In an embodiment of the present application, the charge-discharge power determination module 22 is specifically used for: Input the load prediction value of the power grid in the future time period and the current remaining power of the energy storage element into the reinforcement learning network to obtain the target charge-discharge power of the energy storage element; Wherein, the action space of the reinforcement learning network is constructed based on the theoretical value range of the charge-discharge power, and each action in the action space includes the candidate value of the charge-discharge power of the energy storage element in each of the next N time periods; The calculation process of the reward function of the reinforcement learning network includes: Determine the power grid load fluctuation value based on the N load prediction values of the power grid in the next N time periods; Based on the charge and discharge power candidate values of the energy storage element in each of the next N time periods and the current remaining power of the energy storage element, determine N predicted remaining power values of the energy storage element in the next N time periods; Based on the N predicted remaining power values of the energy storage element in the next N time periods, determine the state of charge evaluation value of the energy storage element; Based on the grid load fluctuation value and the state of charge evaluation value of the energy storage element, calculate the reward function of the reinforcement learning network.

[0068] In an embodiment of the present application, the charge and discharge power determination module 22 is further specifically configured to: Obtain the target load curve of the power grid in the next N time periods; Based on the N load prediction values of the power grid in the next N time periods and the charge and discharge power candidate values of the energy storage element in each of the next N time periods, determine the N net load prediction values of the power grid in the next N time periods; wherein, the net load prediction value is the load prediction value of the power grid itself; Compare the N net load prediction values of the power grid in the next N time periods with the target load curve of the power grid in the next N time periods to obtain the grid load fluctuation value.

[0069] In an embodiment of the present application, the charge and discharge power determination module 22 is further specifically configured to: Obtain the charge and discharge safety threshold of the energy storage element; Compare the N predicted remaining power values of the energy storage element in the next N time periods with the charge and discharge safety threshold to determine the charge and discharge risk evaluation value of the energy storage element in the next N time periods; Based on the N predicted remaining power values of the energy storage element in the next N time periods, determine the change rate of the remaining power between two adjacent time periods to obtain N - 1 remaining power change rates, and based on the N - 1 remaining power change rates, determine the fluctuation value of the remaining power; Perform a weighted sum of the charge and discharge risk evaluation value of the energy storage element in the next N time periods and the fluctuation value of the remaining power to obtain the state of charge evaluation value of the energy storage element.

[0070] In an embodiment of the present application, the charge and discharge power determination module 22 is further specifically configured to: Based on the proportion of new energy power generation in the power grid, determine the first weight of the grid load fluctuation value; wherein, when the proportion of renewable energy power generation in the power grid is greater than the first threshold, the corresponding first weight is greater than the first weight corresponding to when the proportion of renewable energy power generation in the power grid is less than or equal to the first threshold; Based on the first weight, determine the second weight of the state of charge evaluation value of the energy storage element; Perform a weighted sum of the grid load fluctuation value and the state of charge evaluation value of the energy storage element based on the first weight and the second weight to obtain the first summation result; Determine the reciprocal of the first summation result as the reward function of the reinforcement learning network.

[0071] In an embodiment of the present application, the charge and discharge power determination module 22 is specifically further configured to: Input the load prediction value of the power grid in a future period and the current remaining power of the energy storage element into the reinforcement learning network to obtain a target action; Determine the charge and discharge power candidate value of the target period among the N charge and discharge power candidate values included in the target action as the target charge and discharge power of the energy storage element, where the target period is the period closest to the current period within the N periods.

[0072] In an embodiment of the present application, the charge and discharge power determination module 22 is specifically further configured to: Obtain an initial value range; Determine a first adjustment coefficient based on the temperature of the energy storage element; Determine a second adjustment coefficient based on the remaining life of the energy storage element; Adjust the initial value range based on the minimum adjustment coefficient among the first adjustment coefficient and the second adjustment coefficient to obtain the theoretical value range of the charge and discharge power.

[0073] In an embodiment of the present application, the power distribution cabinet 112 may include a controller 300, an energy storage converter, a protection module, etc. Refer to Figure 4 , Figure 4 This is a schematic block diagram of the controller provided in an embodiment of the present application. As Figure 4 shown, the controller 300 in this embodiment may include: one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The above-mentioned processors 301, input devices 302, output devices 303, and memories 304 communicate with each other through a communication bus 305. The memory 304 is used to store computer programs, and the computer programs include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. Among them, the processor 301 is configured to call the program instructions to execute the functions of each module / unit in the above-mentioned device embodiments, such as Figure 3 shown, the functions of the power grid load prediction module 21, the charge and discharge power determination module 22, and the control module 23.

[0074] It should be understood that in the embodiments of the present application, the so-called processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0075] The input device 302 may include a touchpad, a fingerprint acquisition sensor (for acquiring the fingerprint information and the direction information of the fingerprint of the user), a microphone, etc., and the output device 303 may include a display (such as an LCD), a speaker, etc.

[0076] The memory 304 may include a read-only memory and a random access memory, and provide instructions and data to the processor 301. A part of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store preset data such as an initial value range, a target load curve, etc.

[0077] In specific implementation, the processor 301, the input device 302, and the output device 303 described in the embodiments of the present application may implement the implementation manners described in the intelligent control method for the power distribution cabinet provided in the embodiments of the present application, or may also implement the implementation manners of the power distribution cabinet described in the embodiments of the present application, which will not be elaborated herein.

[0078] In another embodiment of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and the computer program includes program instructions. When the program instructions are executed by a processor, all or part of the processes in the methods of the above embodiments are implemented. It can also be completed by instructing relevant hardware through the computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0079] The computer-readable storage medium can be the internal storage unit of the power distribution cabinet in any of the foregoing embodiments, such as the hard disk or memory of the power distribution cabinet. The computer-readable storage medium can also be an external storage device of the power distribution cabinet, such as a plug-in hard disk equipped on the power distribution cabinet, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the computer-readable storage medium can also include both the internal storage unit and the external storage device of the power distribution cabinet. The computer-readable storage medium is used to store the computer program and other programs and data required by the power distribution cabinet. The computer-readable storage medium can also be used to temporarily store the data that has been output or will be output.

[0080] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components and steps of the examples have been generally described according to their functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0081] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described power distribution cabinet and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0082] In several embodiments provided by the present application, it should be understood that the disclosed power distribution cabinet and method can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods. For example, modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed coupling or direct coupling or communication connection between each other can be an indirect coupling or communication connection through some interfaces or units, or can also be an electrical, mechanical or other form of connection.

[0083] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiments of the present application.

[0084] In addition, each functional module in various embodiments of the present application can be integrated in a processing module, or each module can exist physically alone, or two or more modules can be integrated in one module. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules.

[0085] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various equivalent modifications or replacements, and these modifications or replacements should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. An intelligent control method for a power distribution cabinet, which is applied to an energy storage system. The energy storage system includes a power distribution cabinet and energy storage elements. The energy storage elements are connected in parallel with the power grid through the power distribution cabinet, and is characterized in that, The method is executed by the power distribution cabinet, and the method includes: Determining a load prediction value of the power grid in a future period based on historical load data of the power grid; the load prediction value of the power grid in the future period includes N load prediction values of the power grid in the next N periods; Determining a target charge-discharge power of the energy storage element based on the load prediction value of the power grid in the future period and the current remaining power of the energy storage element; Controlling the energy storage element to charge or discharge the power grid according to the target charge-discharge power based on the target charge-discharge power; Wherein, determining the target charge-discharge power of the energy storage element based on the load prediction value of the power grid in the future period and the current remaining power of the energy storage element includes: Inputting the load prediction value of the power grid in the future period and the current remaining power of the energy storage element into a reinforcement learning network to obtain the target charge-discharge power of the energy storage element; Wherein, the action space of the reinforcement learning network is constructed based on the theoretical value range of the charge-discharge power, and each action in the action space includes a charge-discharge power candidate value of the energy storage element in each of the next N periods.

2. The intelligent control method for the power distribution cabinet according to claim 1, wherein The calculation process of the reward function of the reinforcement learning network includes: Determining a power grid load fluctuation value based on N load prediction values of the power grid in the next N periods; Determining N remaining power prediction values of the energy storage element in the next N periods based on the charge-discharge power candidate values of the energy storage element in each of the next N periods and the current remaining power of the energy storage element; Determining a state of charge evaluation value of the energy storage element based on N remaining power prediction values of the energy storage element in the next N periods; Calculating a reward function of the reinforcement learning network based on the power grid load fluctuation value and the state of charge evaluation value of the energy storage element.

3. The intelligent control method for the power distribution cabinet according to claim 2, characterized in that, The determining the power grid load fluctuation value based on N load prediction values of the power grid in the next N periods includes: Obtaining a target load curve of the power grid in the next N periods; Determining N net load prediction values of the power grid in the next N periods based on N load prediction values of the power grid in the next N periods and the charge-discharge power candidate values of the energy storage element in each of the next N periods; wherein, the net load prediction value is the load prediction value of the power grid itself; Comparing the N net load prediction values of the power grid in the next N periods with the target load curve of the power grid in the next N periods to obtain the power grid load fluctuation value.

4. The intelligent control method for the power distribution cabinet according to claim 2, wherein The determining the state of charge evaluation value of the energy storage element based on N remaining power prediction values of the energy storage element in the next N periods includes: Obtaining a charge-discharge safety threshold of the energy storage element; Comparing the N remaining power prediction values of the energy storage element in the next N periods with the charge-discharge safety threshold to determine a charge-discharge risk evaluation value of the energy storage element in the next N periods; Determining a rate of change of the remaining power between adjacent two periods based on N remaining power prediction values of the energy storage element in the next N periods to obtain N-1 rates of change of the remaining power, and determining a fluctuation value of the remaining power based on the N-1 rates of change of the remaining power; Performing a weighted sum of the charge-discharge risk evaluation value of the energy storage element in the next N periods and the fluctuation value of the remaining power to obtain the state of charge evaluation value of the energy storage element.

5. The intelligent control method for the power distribution cabinet according to claim 2, wherein, Calculating the reward function of the reinforcement learning network based on the grid load fluctuation value and the state of charge evaluation value of the energy storage element, including: Determining a first weight of the grid load fluctuation value based on the proportion of new energy power generation in the grid; wherein, when the proportion of renewable energy power generation in the grid is greater than a first threshold, the corresponding first weight is greater than when the proportion of renewable energy power generation in the grid is less than or equal to the first threshold; Determining a second weight of the state of charge evaluation value of the energy storage element based on the first weight; Performing weighted summation on the grid load fluctuation value and the state of charge evaluation value of the energy storage element based on the first weight and the second weight to obtain a first summation result; Determining the reciprocal of the first summation result as the reward function of the reinforcement learning network.

6. The intelligent control method for the power distribution cabinet according to claim 1, characterized in that Inputting the load prediction value of the grid in a future period and the current remaining power of the energy storage element into the reinforcement learning network to obtain the target charge and discharge power of the energy storage element, including: Inputting the load prediction value of the grid in a future period and the current remaining power of the energy storage element into the reinforcement learning network to obtain a target action; Determining the charge and discharge power candidate value in the target period among the N charge and discharge power candidate values included in the target action as the target charge and discharge power of the energy storage element, where the target period is the period closest to the current period within the N periods.

7. The intelligent control method of the power distribution cabinet according to claim 2, wherein The determination method of the theoretical value range of the charge and discharge power, including: Obtaining an initial value range; Determining a first adjustment coefficient based on the temperature of the energy storage element; Determining a second adjustment coefficient based on the remaining life of the energy storage element; Adjusting the initial value range based on the minimum adjustment coefficient among the first adjustment coefficient and the second adjustment coefficient to obtain the theoretical value range of the charge and discharge power.

8. An intelligent control system for a power distribution cabinet is provided in the power distribution cabinet. The power distribution cabinet is included in an energy storage system, and the energy storage system further includes energy storage elements. The energy storage elements are connected in parallel with the power grid through the power distribution cabinet. It is characterized in that, The intelligent control system of the distribution cabinet includes: A grid load prediction module for determining the load prediction value of the grid in a future period based on the historical load data of the grid; the load prediction value of the grid in a future period includes N load prediction values of the grid in N future periods; A charge and discharge power determination module for determining the target charge and discharge power of the energy storage element based on the load prediction value of the grid in a future period and the current remaining power of the energy storage element; A control module for controlling the energy storage element to charge or discharge the grid according to the target charge and discharge power based on the target charge and discharge power; The charge and discharge power determination module is specifically used for: Inputting the load prediction value of the grid in a future period and the current remaining power of the energy storage element into the reinforcement learning network to obtain the target charge and discharge power of the energy storage element; Wherein, the action space of the reinforcement learning network is constructed based on the theoretical value range of the charge and discharge power, and each action in the action space includes the charge and discharge power candidate values of the energy storage element in each of the N future periods.

9. A power distribution cabinet, comprising a controller, the controller including a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images