Energy storage configuration optimization method and system based on deep learning, equipment and medium

Abstract

The invention discloses an energy storage configuration optimization method, system and device based on deep learning and a medium, and relates to the technical field of energy storage optimization, and the method comprises the steps: carrying out the charging test and discharging test of an energy storage device, and obtaining a charging optimization parameter and a discharging optimization parameter; performing primary optimization on the energy storage equipment; establishing a deep learning model and performing training, and performing real-time optimization on the energy storage equipment by using the trained deep training model; the method is used for solving the problems that an existing energy storage configuration optimization method cannot effectively optimize the charging and discharging efficiency selection of energy storage equipment, and when different electric energy is left in a battery of the energy storage equipment, the optimal charging rate or discharging rate cannot be used, so that the energy storage equipment cannot be effectively optimized. Therefore, the lower limit of the charging and discharging efficiency of the energy storage equipment is relatively low during charging and discharging.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage optimization, and specifically to an energy storage configuration optimization method, system, device and medium based on deep learning. Background Art

[0002] Energy storage configuration includes battery energy storage, mechanical energy storage, chemical energy storage, superconducting energy storage, etc. Energy storage configuration optimization refers to a technical means of optimizing and adjusting parameters such as the battery capacity, energy storage efficiency, and output power of an energy storage system to improve the overall efficiency and economy of the energy storage system; specifically, energy storage configuration optimization uses means such as software simulation and experiments to achieve the best energy storage efficiency and economy according to the grid conditions and the performance parameters of the energy storage system.

[0003] Existing methods for energy storage configuration optimization usually establish an analysis model based on the electricity charge calculation rules of users, and obtain the return on investment of different users through the model, so as to obtain a suitable energy storage configuration that meets the users, so as to achieve the purpose of reducing operation losses while meeting the actual energy storage needs of users. Although this improved method can reduce the user cost through dispatching energy storage, there is a lack of effective optimization in the selection of the charge and discharge efficiency of energy storage devices. This will lead to that even if the energy storage scheduling cost of users is reduced by establishing a model, due to the fact that when the batteries of energy storage devices have different remaining electric energies, the optimal charge rate or discharge rate cannot be used, resulting in a lower lower limit of the charge and discharge efficiency when the energy storage devices are charging and discharging, and further causing a lower energy storage efficiency and affecting the energy storage scheduling of users. For example, in the patent application with the publication number CN110717259A, an energy storage configuration and operation optimization method for the user side is disclosed. This solution constructs an energy storage configuration optimization model that combines demand defense and peak shaving and valley filling. For different users, it can use the return on investment to evaluate and select the most suitable energy storage configuration, thereby reducing the user's electricity cost. Other improvements in energy storage configuration optimization usually focus on improving the lifespan of energy storage devices. This improved method still does not effectively optimize the selection of the charge and discharge efficiency of energy storage devices. Even if the charge and discharge power is considered based on the total revenue and cost, due to the fact that when the batteries of energy storage devices have different remaining electric energies, the optimal charge rate or discharge rate cannot be used, resulting in a lower lower limit of the charge and discharge efficiency when the energy storage devices are charging and discharging, and further causing a lower energy storage efficiency and a higher lower limit of the charge and discharge cost. In view of this, it is necessary to improve the existing energy storage configuration optimization methods. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems in the prior art to some extent. By providing an optimization method, system, device and medium for energy storage configuration based on deep learning, it is used to solve the problem that the existing optimization methods for energy storage configuration cannot effectively optimize the selection of the charging and discharging efficiency of energy storage devices. When the batteries of energy storage devices have different remaining electric energies, the optimal charging rate or discharging rate cannot be used, resulting in a lower lower limit of the charging and discharging efficiency during the charging and discharging of energy storage devices, thereby causing a lower energy storage efficiency, affecting the energy storage scheduling of users, and a higher lower limit of the charging and discharging costs.

[0005] To achieve the above object, in the first aspect, the present application provides an optimization method for energy storage configuration based on deep learning, including the following steps: Obtain multiple charging rates capable of charging the energy storage device, and perform a charging test on the energy storage device based on the multiple charging rates; obtain the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test; Establish a discharging test based on the acquisition method of the charging optimization parameters, and perform a discharging test on the energy storage device based on all discharging rates capable of discharging the energy storage unit. Obtain the discharging optimization parameters of each energy storage unit in the energy storage device based on the results of the discharging test; Perform an initial optimization on the configuration of the energy storage device based on the charging optimization parameters and the discharging optimization parameters; Establish a deep learning model, and train the deep learning model based on the acquisition of the charging optimization parameters and the discharging optimization parameters. Use the trained deep training model to perform real-time optimization on the energy storage device.

[0006] Further, performing a charging test on the energy storage device based on multiple charging rates; obtaining the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test includes: Sequentially denote each independent energy storage unit in the energy storage device as energy storage unit CN1 to energy storage unit CN v according to the electric energy that can be stored from small to large; z For any energy storage unit CN z , denote the maximum electric energy that the energy storage unit CN can store as the total electric energy, where z is a positive integer less than or equal to v and greater than or equal to 1; z Perform a charging test on the energy storage unit CN z . The charging test is: sequentially denote the charging rates capable of charging the energy storage unit CN n as chargeable rate CS1 to chargeable rate CS

[0007] Further, charge the energy storage device based on multiple charging rates; obtaining the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test includes: Empty the electric energy in the energy storage unit CN z and charge the energy storage unit CN with the chargeable rate CS1 z ; establish a plane rectangular coordinate system, denoted as the charging test coordinate system, where the unit of the X-axis of the charging test coordinate system is time and the unit of the Y-axis is electric energy; when charging starts, based on the electric energy in the energy storage unit CN z draw a corresponding curve in the charging test coordinate system until the energy storage unit CN z is fully charged, and record the obtained curve as the charging energy curve CQ1; Empty the electric energy in the energy storage unit CN z and charge the energy storage unit CN with the chargeable rate CS2 z ; and obtain the charging energy curve CQ2 based on the obtaining method of the charging energy curve CQ1, and so on until the charging energy curve CQ n is obtained.

[0008] Further, charge the energy storage device based on multiple charging rates; obtaining the charging optimization parameters of each energy storage unit in the energy storage device further includes: Put all the charging energy curves CQ in the same charging test coordinate system, and record the point with the largest abscissa among the rightmost points of all the charging energy curves CQ as the charging extreme right point; use a straight line parallel to the Y-axis to evenly divide the line segment from the coordinate origin to X = the charging extreme right point into k parts, and sequentially denote each part from left to right as the charging horizontal interval CH1 to the charging horizontal interval CH k ; for any charging horizontal interval CH j , record the point with the largest slope among all the curves in the charging horizontal interval CH j as the interval charging peak point of the charging horizontal interval CH j ; obtain the interval charging peak points of all the charging horizontal intervals CH, where j is a positive integer less than or equal to k and greater than or equal to 1; Use a straight line parallel to the X-axis to evenly divide the line segment from the coordinate origin to Y = the total electric energy into k parts, and sequentially denote each part from bottom to top as the charging vertical interval CZ1 to the charging vertical interval CZ k ; for any charging vertical interval CZ j , when the number of interval charging peak points in all the curves in the charging vertical interval CZ j is greater than 1, record the interval charging peak point with the largest slope among all the interval charging peak points as the preferred charging peak point; when the number of interval charging peak points in all the curves in the charging vertical interval CZ j is equal to 1, record the charging vertical interval CZj The interval charging peak point within it is denoted as the preferred charging peak point; when the number of interval charging peak points among all the curves within the charging longitudinal interval CZ j is equal to 0, the point with the maximum slope among all the curves within the charging longitudinal interval CZ j is denoted as the preferred charging peak point; Obtain the preferred charging peak points corresponding to all the charging longitudinal intervals CZ; Denote all the charging longitudinal intervals CZ and their preferred charging peak points as the charging optimization parameters of the energy storage unit CN z of the energy storage unit CN

[0009] Furthermore, establish a discharge test based on the acquisition method of the charging optimization parameters, and conduct a discharge test on the energy storage device at all discharge rates capable of discharging the energy storage unit. The discharge optimization parameters of each energy storage unit in the energy storage device obtained based on the results of the discharge test include: Conduct a discharge test on the energy storage unit CN z The discharge test is as follows: Based on the charging test, obtain all the discharge rates capable of discharging the energy storage unit CN. After filling the electric energy in the energy storage unit CN, sequentially use all the discharge rates to discharge the charging unit CN, and establish a rectangular coordinate system to obtain the curves corresponding to all the discharge rates; Analyze the curves corresponding to all the discharge rates based on the acquisition method of the charging optimization parameters. Denote the interval obtained by the straight line parallel to the X-axis as the discharge longitudinal interval FZ, and denote the point obtained based on the preferred charging peak point as the preferred discharge peak point; Denote all the discharge longitudinal intervals FZ and their preferred discharge peak points as the discharge optimization parameters of the energy storage unit CN z of the energy storage unit CN; Obtain the charging optimization parameters and discharge optimization parameters of all the energy storage units CN.

[0010] Furthermore, conduct a primary optimization of the configuration of the energy storage device based on the charging optimization parameters and the discharge optimization parameters, including: The primary optimization includes: When charging the energy storage battery, obtain the electric energy of each energy storage unit CN in the energy storage battery in real time; For any unfilled energy storage unit CN z , obtain the charging optimization parameters of the charging unit CN z , Denote the charging longitudinal interval CZ where the electric energy in the energy storage unit CN z is located as the preferred charging interval, and denote the chargeable rate CS corresponding to the curve where the preferred charging peak point of the energy storage unit CN z in the preferred charging interval is located as the preferred charging rate, and use the preferred charging rate to charge the energy storage unit CN z ; When discharging the energy storage battery, obtain the electric energy of each energy storage unit CN in the energy storage battery in real time; For any unfilled energy storage unit CNz , obtain the discharge optimization parameters of the charging unit CN z , and record the longitudinal discharge interval FZ of the electric energy in the energy storage unit CN z as the preferred discharge interval. Record the discharge rate corresponding to the curve where the preferred discharge peak point of the energy storage unit CN z in the preferred discharge interval is located as the preferred discharge rate, and use the preferred discharge rate to discharge the energy storage unit CN z .

[0011] Further, establish a deep learning model, and train the deep learning model based on the acquisition of the charging optimization parameters and the discharge optimization parameters. Using the trained deep learning model to perform real-time optimization on the energy storage device includes: Establish a deep learning model, store the charge and discharge data of the energy storage battery after the initial optimization in the deep learning model. The unit at the input end of the deep learning model is used to input the type of electricity consumption and the remaining electric energy, and the unit at the output end of the deep learning model is used to output the charge and discharge rate. Among them, the type of electricity consumption includes charging and discharging, and the charge and discharge rate includes the chargeable rate and the discharge rate; Based on the charge and discharge data of the energy storage battery stored in the deep learning model, perform real-time optimization on the charging optimization parameters and the discharge optimization parameters of each energy storage unit CN. The optimization method is: When the energy storage unit CN is charging or discharging, input the type of electricity consumption and the remaining electric energy of the energy storage unit CN into the deep learning model. Based on the simulation of all chargeable rates or discharge rates, obtain the fastest chargeable rate or discharge rate when the energy storage unit CN is fully charged or emptied from the remaining electric energy, and output it by the unit at the output end; When the output result of the deep learning model is the same as the preferred charging rate or the preferred discharge rate obtained by using the initial optimization, do not optimize the charging optimization parameters and the discharge optimization parameters of the energy storage unit CN; When the output result of the deep learning model is different from the preferred charging rate or the preferred discharge rate obtained by using the initial optimization, use the optimization selection algorithm to obtain the optimization judgment value. The optimization selection algorithm is: , where G is the optimization judgment value, f0 is the output result of the deep learning model. When the output result of the deep learning is the chargeable rate, g is the number of preferred charging rates obtained by the initial optimization of the energy storage unit CN, and f i is the i-th preferred charging rate obtained by the initial optimization of the energy storage unit CN; when the output result of the deep learning is the discharge rate, g is the number of preferred discharge rates obtained by the initial optimization of the energy storage unit CN, and f i is the i-th preferred discharge rate obtained by the initial optimization of the energy storage unit CN; When the optimization judgment value is greater than 0, the output result of the deep learning model is used to replace the preferred charging rate or the preferred discharging rate of the energy storage unit CN; When the optimization judgment value is less than or equal to 0, the charging optimization parameters and the discharging optimization parameters of the energy storage unit CN are not optimized.

[0012] In a second aspect, the present application further provides an energy storage configuration optimization system based on deep learning, including a charging optimization analysis module, a discharging optimization analysis module, and a deep optimization module; The charging optimization analysis module is configured to obtain a plurality of charging rates capable of charging the energy storage device, and perform a charging test on the energy storage device based on the plurality of charging rates; obtain the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test; The discharging optimization analysis module is configured to establish a discharging test based on the obtaining manner of the charging optimization parameters, and perform a discharging test on the energy storage device based on all discharging rates capable of discharging the energy storage unit, and obtain the discharging optimization parameters of each energy storage unit in the energy storage device based on the results of the discharging test; Perform an initial optimization on the configuration of the energy storage device based on the charging optimization parameters and the discharging optimization parameters; The deep optimization module is configured to establish a deep learning model, train the deep learning model based on the obtaining of the charging optimization parameters and the discharging optimization parameters, and perform real-time optimization on the energy storage device using the trained deep training model.

[0013] In a third aspect, the present application provides an electronic device, including a processor and a memory, where the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in the above method are run.

[0014] In a fourth aspect, the present application provides a storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps in the above method are run.

[0015] Advantages of the present invention: The present application first obtains a plurality of charging rates capable of charging the energy storage device, and performs a charging test on the energy storage device based on the plurality of charging rates; obtains the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test; then performs a discharging test based on the obtaining manner of the charging optimization parameters, and obtains the discharging optimization parameters of each energy storage unit in the energy storage device based on the results of the discharging test. The advantage of this is that by performing the charging test and the discharging test, and obtaining the charging optimization parameters and the discharging optimization parameters, the optimal charging rate and discharging can be obtained when the battery of the energy storage device has different remaining electric energies, so that in subsequent management, even considering the charging and discharging costs and user scheduling, the charging and discharging efficiency can still be ensured to be high, the lower limit of the charging and discharging costs can be reduced, and the energy storage scheduling efficiency of the user can be improved; This application also initially optimizes the configuration of the energy storage device based on the charging optimization parameters and the discharging optimization parameters; finally, a deep learning model is established, and the deep learning model is trained based on the acquisition of the charging optimization parameters and the discharging optimization parameters, and the trained deep training model is used to perform real-time optimization on the energy storage device. The advantage of this is that by initially optimizing the energy storage device and establishing a deep learning model to update the charging optimization parameters and the discharging optimization parameters in real time, while applying the charging optimization parameters and the discharging optimization parameters, the charging optimization parameters and the discharging optimization parameters can be updated in real time based on the real-time state of the energy storage device to ensure that the charging optimization parameters and the discharging optimization parameters used are both the optimal parameters that conform to the current operating state of the energy storage device. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a schematic block diagram of the system of the present invention; Figure 2 is a flowchart of the steps of the method of the present invention; Figure 3 is a schematic diagram of the charging test coordinate system of the present invention; Figure 4 is a schematic structural diagram of the deep learning model of the present invention; Figure 5 is a schematic structural diagram of the electronic device of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment 1, please refer to Figure 1 As shown, this application provides an energy storage configuration optimization system based on deep learning, including a charging optimization analysis module, a discharging optimization analysis module, and a deep optimization module; The charging optimization analysis module is used to obtain a plurality of charging rates capable of charging the energy storage device, and perform a charging test on the energy storage device based on the plurality of charging rates; obtain the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test; The charging optimization analysis module includes a charging parameter analysis unit, and the charging parameter analysis unit is configured with a charging parameter analysis strategy, and the charging parameter analysis strategy includes: Each independent energy storage unit in the energy storage device is sequentially recorded as energy storage unit CN1 to energy storage unit CN according to the electrical energy that can be stored from small to largev ; For any energy storage unit CN z , denote the maximum electric energy that the energy storage unit CN z can store as the total electric energy, where z is a positive integer less than or equal to v and greater than or equal to 1; In the specific implementation process, for example, in actual applications, if the energy storage device is a battery pack formed by connecting multiple batteries, then the energy storage unit can be multiple independent batteries in the energy storage device; Conduct a charging test on the energy storage unit CN z . The charging test is as follows: Denote the charging rates that can charge the energy storage unit CN z from small to large as chargeable rate CS1 to chargeable rate CS n ; In the specific implementation process, for example, in a data processing, if the charging rates for charging the energy storage unit CN are 0.25C, 0.5C, 1C, and 2C respectively, then the chargeable rates CS1 to CS4 can be denoted as 0.25C, 0.5C, 1C, and 2C in sequence; By obtaining the chargeable rate CS, various charging rates of the energy storage unit CN can be obtained, thus providing data support for subsequent analysis; Empty the electric energy in the energy storage unit CN z , and use the chargeable rate CS1 to charge the energy storage unit CN z . Establish a plane rectangular coordinate system, denoted as the charging test coordinate system. Among them, the unit of the X-axis of the charging test coordinate system is time, and the unit of the Y-axis is electric energy; When the charging starts, draw the corresponding curve in the charging test coordinate system based on the electric energy in the energy storage unit CN z , until the energy storage unit CN z is fully charged, and denote the obtained curve as the charging energy curve CQ1; Empty the electric energy in the energy storage unit CN z , use the chargeable rate CS2 to charge the energy storage unit CN z , and obtain the charging energy curve CQ2 based on the acquisition method of the charging energy curve CQ1, and so on, until the charging energy curve CQ n is obtained; Put all the charging energy curves CQ in the same charging test coordinate system, and denote the point with the largest abscissa among the rightmost points of all the charging energy curves CQ as the charging extreme right point; Use a straight line parallel to the Y-axis to evenly divide the line segment from the coordinate origin to X = charging extreme right point into k parts, and denote each part from left to right as charging horizontal interval CH1 to charging horizontal interval CH k ; For any charging horizontal interval CH j , denote the point with the largest slope among all the curves in the charging horizontal interval CH j as the charging horizontal interval CH jInterval charging peak points; obtain the interval charging peak points of all charging horizontal intervals CH, where j is a positive integer less than or equal to k and greater than or equal to 1; In the specific implementation process, the value of k can be set according to the actually analyzable data volume. The larger the value of k, the more data needs to be processed, and the more accurate the analysis data obtained. In this embodiment, the value of k is set to 5 for analysis; For example, in a data processing, the obtained charging test coordinate system is as Figure 3 shown, where the curves CQ1 to CQ4 are the charging curves CQ1 to CQ4 respectively. Through analysis, it can be obtained that the point JY is the right charging point, and the interval charging peak points in the charging horizontal intervals CH1 to CH5 are the points FD1 to FD5; by obtaining the interval charging peak points, it is possible to obtain the points with the fastest charging rate in different time intervals after the charging of the energy storage unit CN starts, which helps to provide data support for the screening of the preferred charging peak points in subsequent analysis; at the same time, considering that the actual charging rate of the energy storage unit CN does not match the expectation due to the influence of the charging duration, therefore, by obtaining the points with the fastest charging rate in different time intervals, it is possible to analyze in combination with the actual charging situation of the energy storage unit CN to ensure that the obtained data can fit the actual application scenario; Use a straight line parallel to the X-axis to divide the origin of coordinates to Y = total electric energy into k equal parts, and sequentially record each part from bottom to top as the charging vertical intervals CZ1 to CZ k ; For any charging vertical interval CZ j , when the number of interval charging peak points in all curves within the charging vertical interval CZ j is greater than 1, record the interval charging peak point with the largest slope among all interval charging peak points as the preferred charging peak point; when the number of interval charging peak points in all curves within the charging vertical interval CZ j is equal to 1, record the interval charging peak point within the charging vertical interval CZ j as the preferred charging peak point; when the number of interval charging peak points in all curves within the charging vertical interval CZ j is equal to 0, record the point with the largest slope among all curves within the charging vertical interval CZ j as the preferred charging peak point; In this embodiment, the purpose of obtaining the interval charging peak points in multiple intervals is to uniformly obtain the fastest charging rate in each charging time period during the charging process. Therefore, whether the end of the interval is steep does not affect the acquisition of the interval charging peak points. At the same time, the screening of the optimal charging curve is also obtained through the interval charging peak points or the points with the largest slope in the charging longitudinal interval. Therefore, there is no need to worry about the influence of the steep end of a certain charging transverse interval on the acquisition of the optimal charging curve. If the curve slope is fixed in a certain interval, the position of the charging peak point can be set by itself, or the charging transverse interval with a fixed curve slope can be calibrated, and when analyzing the charging longitudinal interval subsequently, the points on the curve within the calibrated charging transverse interval are all recorded as interval charging peak points and the optimal charging peak points are screened; Different charging curves of the same energy storage unit CN are all obtained at different charging rates. "The charging rates that can charge the energy storage unit CN z are sequentially recorded as chargeable rates CS1 to chargeable rate CS n from small to large." "When charging starts, based on the electric energy in the energy storage unit CN z a corresponding curve is plotted in the charging test coordinate system until the energy storage unit CN z is fully charged, and the obtained curve is recorded as the charging curve CQ1; the electric energy in the energy storage unit CN z is emptied, and the energy storage unit CN z is charged using the chargeable rate CS2, and the charging curve CQ2 is obtained based on the acquisition method of the charging curve CQ1, and so on until the charging curve CQ n is obtained." Since for different charging curves, the conditional factors other than the charging rate, such as the self-configuration environment of the energy storage unit or the external environment, are the same, different charging curves are comparable.

[0019] In the specific implementation process, by obtaining the optimal charging peak points, the optimal charging rate of the energy storage unit CN at different remaining electric energies can be obtained, which helps to use the fastest charging rate when charging the remaining unused energy storage units CN, so as to improve the upper limit of the charging efficiency; Obtain the optimal charging peak points corresponding to all charging longitudinal intervals CZ; record all charging longitudinal intervals CZ and their optimal charging peak points as the charging optimization parameters of the energy storage unit CN z .

[0020] The discharge optimization analysis module is used to establish a discharge test based on the acquisition method of the charging optimization parameters, and discharge the energy storage device based on all discharge rates that can discharge the energy storage unit, and obtain the discharge optimization parameters of each energy storage unit in the energy storage device based on the results of the discharge test; Perform an initial optimization of the configuration of the energy storage device based on the charging optimization parameters and the discharge optimization parameters; The discharge optimization analysis module includes a discharge parameter analysis unit, and the discharge parameter analysis unit is configured with a discharge parameter analysis strategy, and the discharge parameter analysis strategy includes: Perform a discharge test on the energy storage unit CN z The discharge test is as follows: Based on the charge test, obtain all discharge rates at which the energy storage unit CN can be discharged. After the electric energy in the energy storage unit CN is fully charged, use all discharge rates to discharge the charging unit CN in sequence, and establish a rectangular coordinate system to obtain the curves corresponding to all discharge rates; Analyze the curves corresponding to all discharge rates based on the acquisition method of the charge optimization parameters. Denote the interval obtained by the straight line parallel to the X-axis as the discharge longitudinal interval FZ, and denote the point based on the preferred charge peak point as the preferred discharge peak point; Denote all the discharge longitudinal intervals FZ and their preferred discharge peak points as the discharge optimization parameters of the energy storage unit CN z ; In the specific implementation process, refer to the method of using multiple chargeable rates CS in the charge test and obtaining the charge curve. Based on the discharge rate, obtain the corresponding curve, and put all the discharge rates into the same coordinate system. Use the straight lines parallel to the Y-axis and the X-axis to analyze in sequence based on the analysis method of the charge test, and obtain the discharge longitudinal interval FZ and its preferred discharge peak point corresponding to the discharge test based on the analysis results; The purpose is to obtain the optimal discharge rate of the energy storage unit CN when the remaining electric energy is different, which helps to use the fastest discharge rate when discharging the remaining unused energy storage unit CN to improve the upper limit of the discharge efficiency; Obtain the charge optimization parameters and discharge optimization parameters of all energy storage units CN; The initial optimization includes: When charging the energy storage battery, obtain the electric energy of each energy storage unit CN in the energy storage battery in real time; For any unfilled energy storage unit CN z , obtain the charge optimization parameters of the charging unit CN z , Denote the charging longitudinal interval CZ where the electric energy in the energy storage unit CN z is located as the preferred charging interval, and denote the chargeable rate CS corresponding to the curve where the preferred charge peak point of the energy storage unit CN z in the preferred charging interval is located as the preferred charging rate, and use the preferred charging rate to charge the energy storage unit CN z ; When discharging the energy storage battery, obtain the electric energy of each energy storage unit CN in the energy storage battery in real time; For any unfilled energy storage unit CN z , obtain the discharge optimization parameters of the charging unit CN z , Denote the energy storage unit CN zThe longitudinal discharge interval FZ of the electrical energy in it is denoted as the preferred discharge interval, and the energy storage unit CN in the preferred discharge interval z The discharge rate corresponding to the curve where the preferred discharge peak point of is located is denoted as the preferred discharge rate, and the energy storage unit CN is discharged using the preferred discharge rate. z Discharge.

[0021] The deep optimization module is used to establish a deep learning model, and based on the acquisition of charging optimization parameters and discharge optimization parameters, train the deep learning model, and use the trained deep training model to optimize the energy storage device in real time; The deep optimization module includes a deep real-time optimization unit, and the deep real-time optimization unit is configured with a deep real-time optimization strategy, and the deep real-time optimization strategy includes: Establish a deep learning model, store the charge and discharge data of the energy storage battery after the first optimization in the deep learning model, the unit at the input end of the deep learning model is used to input the type of electricity consumption and the remaining electrical energy, and the unit at the output end of the deep learning model is used to output the charge and discharge rate. Among them, the type of electricity consumption includes charging and discharging, and the charge and discharge rate includes the chargeable rate and the discharge rate; In the specific implementation process, the structure of the deep learning model in this embodiment can refer to Figure 4 As shown; Based on the charge and discharge data of the energy storage battery stored in the deep learning model, the charging optimization parameters and discharge optimization parameters of each energy storage unit CN are optimized in real time, and the optimization method is: When the energy storage unit CN is charging or discharging, input the type of electricity consumption and the remaining electrical energy of the energy storage unit CN into the deep learning model, and based on the simulation of all chargeable rates or discharge rates, obtain the fastest chargeable rate or discharge rate when the energy storage unit CN is fully charged or emptied from the remaining electrical energy, and output it by the unit at the output end; When the output result of the deep learning model is the same as the preferred charging rate or preferred discharge rate obtained by the first optimization, do not optimize the charging optimization parameters and discharge optimization parameters of the energy storage unit CN; When the output result of the deep learning model is different from the preferred charging rate or preferred discharge rate obtained by the first optimization, use the optimization selection algorithm to obtain the optimization judgment value, and the optimization selection algorithm is , where G is the optimization judgment value, f0 is the output result of the deep learning model. When the output result of the deep learning is the chargeable rate, g is the number of preferred charging rates obtained by the first optimization of the energy storage unit CN, and f i is the i-th preferred charging rate obtained by the first optimization of the energy storage unit CN; when the output result of the deep learning is the discharge rate, g is the number of preferred discharge rates obtained by the first optimization of the energy storage unit CN, and f i is the i-th preferred discharge rate obtained by the first optimization of the energy storage unit CN; In a specific implementation process, for example, during a data processing, the output result of deep learning is that the charge rate is 2C. The number of preferred charge rates obtained by the initial optimization of the energy storage unit CN is 5, and the preferred charge rates obtained by the initial optimization of the energy storage unit CN are successively: 1C, 2C, 2C, 2C, and 1C. Then, through calculation, the optimization judgment value is 2, that is, the optimization judgment value is greater than 0, indicating that the charge rate obtained by the deep learning model is more efficient than the preferred charge rate obtained by the initial optimization when charging the energy storage unit CN. Therefore, 2C can be used to replace the preferred charge rate of the energy storage unit CN to achieve higher-efficiency charging of the energy storage unit CN; When the optimization judgment value is greater than 0, replace the preferred charge rate or the preferred discharge rate of the energy storage unit CN with the output result of the deep learning model; When the optimization judgment value is less than or equal to 0, do not optimize the charging optimization parameters and the discharge optimization parameters of the energy storage unit CN.

[0022] Example 2, please refer to Figure 2 As shown, the present application also provides an energy storage configuration optimization method based on deep learning, including the following steps: Step S1, obtain multiple charge rates capable of charging the energy storage device, and perform a charging test on the energy storage device based on the multiple charge rates; obtain the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test; Step S1 includes: Step S101, sequentially record each independent energy storage unit in the energy storage device as energy storage unit CN1 to energy storage unit CN v according to the electrical energy that can be stored from small to large; for any energy storage unit CN z , record the maximum electrical energy that the energy storage unit CN z can store as the total electrical energy, where z is a positive integer less than or equal to v and greater than or equal to 1; Step S102, perform a charging test on the energy storage unit CN z , and the charging test is: sequentially record the charge rates capable of charging the energy storage unit CN z from small to large as charge rate CS1 to charge rate CS n ; Step S103, empty the electrical energy in the energy storage unit CN z , and use the charge rate CS1 to charge the energy storage unit CN z ; establish a plane rectangular coordinate system, denoted as the charging test coordinate system, where the unit of the X-axis of the charging test coordinate system is time, and the unit of the Y-axis is electrical energy; when charging starts, draw a corresponding curve in the charging test coordinate system based on the electrical energy in the energy storage unit CN z , until the energy storage unit CNz When it is full, the obtained curve is denoted as the charging energy curve CQ1; Step S104, empty the electric energy in the energy storage unit CN z and charge the energy storage unit CN at the chargeable rate CS2, and obtain the charging energy curve CQ2 based on the obtaining method of the charging energy curve CQ1, and so on until the charging energy curve CQ z is obtained; n ; Step S105, place all the charging energy curves CQ in the same charging test coordinate system, and denote the point with the largest abscissa among the rightmost points of all the charging energy curves CQ as the charging extreme right point; use a straight line parallel to the Y-axis to evenly divide the line segment from the coordinate origin to X = the charging extreme right point into k parts, and sequentially denote each part from left to right as the charging horizontal interval CH1 to the charging horizontal interval CH k ; For any charging horizontal interval CH j , denote the point with the largest slope among all the curves in the charging horizontal interval CH j as the interval charging peak point of the charging horizontal interval CH j ; Obtain the interval charging peak points of all the charging horizontal intervals CH, where j is a positive integer less than or equal to k and greater than or equal to 1; Step S106, use a straight line parallel to the X-axis to evenly divide the line segment from the coordinate origin to Y = the total electric energy into k parts, and sequentially denote each part from bottom to top as the charging vertical interval CZ1 to the charging vertical interval CZ k ; For any charging vertical interval CZ j , when the number of interval charging peak points among all the curves in the charging vertical interval CZ j is greater than 1, denote the interval charging peak point with the largest slope among all the interval charging peak points as the preferred charging peak point; when the number of interval charging peak points among all the curves in the charging vertical interval CZ j is equal to 1, denote the interval charging peak point in the charging vertical interval CZ j as the preferred charging peak point; when the number of interval charging peak points among all the curves in the charging vertical interval CZ j is equal to 0, denote the point with the largest slope among all the curves in the charging vertical interval CZ j as the preferred charging peak point; Step S107, obtain the preferred charging peak points corresponding to all the charging vertical intervals CZ; denote all the charging vertical intervals CZ and their preferred charging peak points as the charging optimization parameters of the energy storage unit CN z .

[0023] Step S2: Establish a discharge test based on the acquisition method of the charge optimization parameters, perform a discharge test on the energy storage device at all discharge rates capable of discharging the energy storage unit, and obtain the discharge optimization parameters of each energy storage unit in the energy storage device based on the results of the discharge test; Perform an initial optimization of the configuration of the energy storage device based on the charge optimization parameters and the discharge optimization parameters; Step S2 includes: Step S201, perform a discharge test on the energy storage unit CN z The discharge test is as follows: Based on the charge test, obtain all discharge rates capable of discharging the energy storage unit CN. After the electric energy in the energy storage unit CN is fully charged, sequentially use all discharge rates to discharge the charging unit CN, and establish a rectangular coordinate system to obtain the curves corresponding to all discharge rates; Step S202: Analyze the curves corresponding to all discharge rates based on the acquisition method of the charge optimization parameters. Denote the interval obtained by the straight line parallel to the X-axis as the discharge longitudinal interval FZ, and denote the point based on the preferred charge peak point as the preferred discharge peak point; Denote all the discharge longitudinal intervals FZ and their preferred discharge peak points as the discharge optimization parameters of the energy storage unit CN z ; Step S203: Obtain the charge optimization parameters and the discharge optimization parameters of all energy storage units CN; Step S204: The initial optimization includes: Step S2041, when charging the energy storage battery, obtain the electric energy of each energy storage unit CN in the energy storage battery in real time; For any unfilled energy storage unit CN z , obtain the charge optimization parameters of the charging unit CN z , denote the charge longitudinal interval CZ where the electric energy in the energy storage unit CN z is located as the preferred charge interval, denote the charge rate CS corresponding to the curve where the preferred charge peak point of the energy storage unit CN z is located in the preferred charge interval as the preferred charge rate, and use the preferred charge rate to charge the energy storage unit CN z ; Step S2042, when discharging the energy storage battery, obtain the electric energy of each energy storage unit CN in the energy storage battery in real time; For any unfilled energy storage unit CN z , obtain the discharge optimization parameters of the charging unit CN z , denote the discharge longitudinal interval FZ where the electric energy in the energy storage unit CN z is located as the preferred discharge interval, denote the discharge rate corresponding to the curve where the preferred discharge peak point of the energy storage unit CN z is located in the preferred discharge interval as the preferred discharge rate, and use the preferred discharge rate to discharge the energy storage unit CN z .

[0024] Step S3: Establish a deep learning model, and train the deep learning model based on the obtained charging optimization parameters and discharging optimization parameters, and use the trained deep learning model to perform real-time optimization on the energy storage device.

[0025] Step S3 includes: Step S301: Establish a deep learning model, store the charge and discharge data of the energy storage battery after the initial optimization in the deep learning model. The unit at the input end of the deep learning model is used to input the type of electricity consumption and the remaining electric energy, and the unit at the output end of the deep learning model is used to output the charge and discharge rate. Among them, the type of electricity consumption includes charging and discharging, and the charge and discharge rate includes the chargeable rate and the discharge rate. Step S302: Based on the charge and discharge data of the energy storage battery stored in the deep learning model, perform real-time optimization on the charging optimization parameters and discharging optimization parameters of each energy storage unit CN. The optimization method is as follows: Step S303: When the energy storage unit CN is charging or discharging, input the type of electricity consumption and the remaining electric energy of the energy storage unit CN into the deep learning model. Based on the simulation of all chargeable rates or discharge rates, obtain the fastest chargeable rate or discharge rate when the energy storage unit CN is fully charged or emptied from the remaining electric energy, and output it by the unit at the output end. Step S304: When the output result of the deep learning model is the same as the preferred charging rate or preferred discharge rate obtained by the initial optimization, do not optimize the charging optimization parameters and discharging optimization parameters of the energy storage unit CN. Step S305: When the output result of the deep learning model is different from the preferred charging rate or preferred discharge rate obtained by the initial optimization, use an optimization selection algorithm to obtain an optimization judgment value. The optimization selection algorithm is as follows: , where G is the optimization judgment value, f0 is the output result of the deep learning model. When the output result of the deep learning is the chargeable rate, g is the number of preferred charging rates obtained by the initial optimization of the energy storage unit CN, and f i is the i-th preferred charging rate obtained by the initial optimization of the energy storage unit CN; when the output result of the deep learning is the discharge rate, g is the number of preferred discharge rates obtained by the initial optimization of the energy storage unit CN, and f i is the i-th preferred discharge rate obtained by the initial optimization of the energy storage unit CN; Step S306: When the optimization judgment value is greater than 0, replace the preferred charging rate or preferred discharge rate of the energy storage unit CN with the output result of the deep learning model. Step S307: When the optimization judgment value is less than or equal to 0, do not optimize the charging optimization parameters and discharging optimization parameters of the energy storage unit CN.

[0026] Example 3, please refer to Figure 5 as shown inFigure 5 The structural schematic diagram of an electronic device is exemplified. The electronic device may include: a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete communication with each other through the communication bus. The memory stores computer-readable instructions. The processor can call the instructions in the memory. When the computer-readable instructions are executed by the processor, the steps in the energy storage configuration optimization method based on deep learning are run to achieve the following functions: First, obtain multiple charging rates capable of charging the energy storage device, and conduct a charging test on the energy storage device based on the multiple charging rates; obtain the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test; then conduct a discharging test based on the way of obtaining the charging optimization parameters, and obtain the discharging optimization parameters of each energy storage unit in the energy storage device based on the results of the discharging test; and conduct an initial optimization of the configuration of the energy storage device based on the charging optimization parameters and the discharging optimization parameters; finally, establish a deep learning model, and train the deep learning model based on the acquisition of the charging optimization parameters and the discharging optimization parameters, and use the trained deep learning model to perform real-time optimization on the energy storage device.

[0027] In addition, when the logical instructions in the above-mentioned memory are implemented in the form of software functional units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. And the aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0028] Embodiment 4. The present application also provides a computer-readable storage medium. The present application provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, it runs the steps in the above-mentioned energy storage configuration optimization method based on deep learning to achieve the following functions: First, obtain multiple charging rates capable of charging the energy storage device, and perform a charging test on the energy storage device based on the multiple charging rates; obtain the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test; then perform a discharge test based on the way of obtaining the charging optimization parameters, and obtain the discharge optimization parameters of each energy storage unit in the energy storage device based on the results of the discharge test; and perform an initial optimization of the configuration of the energy storage device based on the charging optimization parameters and the discharge optimization parameters; finally, establish a deep learning model, and train the deep learning model based on the acquisition of the charging optimization parameters and the discharge optimization parameters, and use the trained deep learning model to perform real-time optimization on the energy storage device.

[0029] Through the description of the above embodiments, the embodiments of the present invention can be provided as a method, a system or a computer program product. Based on such an understanding, the above technical solution, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0030] In the embodiments provided by the present application, it should be understood that the disclosed system or method can be implemented in other ways. The above-described embodiments are merely illustrative. For example, the division of modules or units is only a logical function division, and there can be other division methods in actual implementation. For another example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some communication interfaces. The indirect coupling or communication connection of systems, modules, and units can be electrical, mechanical, or other forms.

[0031] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. An energy storage configuration optimization method based on deep learning, characterized in that, It includes the following steps: Obtain multiple charging rates capable of charging the energy storage device, and conduct a charging test on the energy storage device based on the multiple charging rates; obtain the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test; Establish a discharge test based on the way of obtaining the charging optimization parameters, and conduct a discharge test on the energy storage device based on all discharge rates capable of discharging the energy storage unit. Obtain the discharge optimization parameters of each energy storage unit in the energy storage device based on the results of the discharge test; Conduct a primary optimization of the configuration of the energy storage device based on the charging optimization parameters and the discharge optimization parameters; Establish a deep learning model, train the deep learning model based on the acquisition of the charging optimization parameters and the discharge optimization parameters, and use the trained deep learning model to conduct real-time optimization of the energy storage device.

2. The energy storage configuration optimization method based on deep learning according to claim 1, wherein Conduct a charging test on the energy storage device based on multiple charging rates; Obtaining the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test includes: Each independent energy storage unit in the energy storage device is sequentially denoted as energy storage unit CN1 to energy storage unit CN according to the amount of electrical energy that can be stored, from small to large. v ; For any energy storage unit CN z , the maximum electrical energy that the energy storage unit CN z can store is denoted as the total electrical energy, where z is a positive integer less than or equal to v and greater than or equal to 1; Perform a charging test on the energy storage unit CN z The charging test is as follows: The charging rates capable of charging the energy storage unit CN z are sequentially recorded as chargeable rates CS1 to CS n .

3. The energy storage configuration optimization method based on deep learning according to claim 2, wherein, Conduct a charging test on the energy storage device based on multiple charging rates; Obtaining the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test includes: Empty the electrical energy in the energy storage unit CN z and charge the energy storage unit CN at the charge rate CS1 z ; Establish a rectangular coordinate system, denoted as the charge test coordinate system. Among them, the unit of the X-axis of the charge test coordinate system is time, and the unit of the Y-axis is electrical energy; When the charging starts, based on the electrical energy in the energy storage unit CN z draw the corresponding curve in the charge test coordinate system until the energy storage unit CN z is fully charged, and the obtained curve is denoted as the charge energy curve CQ1; Empty the electric energy in the energy storage unit CN z and charge the energy storage unit CN z at the chargeable rate CS2, and obtain the charging curve CQ2 based on the acquisition method of the charging curve CQ1, and so on until the charging curve CQ n is obtained.

4. The energy storage configuration optimization method based on deep learning according to claim 3, characterized in that Conduct a charging test on the energy storage device based on multiple charging rates; Obtaining the charging optimization parameters of each energy storage unit in the energy storage device based on the results of the charging test further includes: Put all the charging curves CQ in the same charging test coordinate system, and mark the point with the largest abscissa among the rightmost points of all the charging curves CQ as the charging extreme right point; use a straight line parallel to the Y-axis to evenly divide the coordinate origin to X = the charging extreme right point into k parts, and sequentially mark each part as the charging horizontal interval CH1 to the charging horizontal interval CH from left to right k ; For any charging horizontal interval CH j , j mark the point with the largest slope among all the curves within the charging horizontal interval CH j as the interval charging peak point of the charging horizontal interval CH ; Obtain the interval charging peak points of all the charging horizontal intervals CH, where j is a positive integer less than or equal to k and greater than or equal to 1; Use a straight line parallel to the X-axis to divide the total electric energy from the coordinate origin to Y into k equal parts, and sequentially label each part from bottom to top as the charging longitudinal interval CZ1 to the charging longitudinal interval CZ k ; for any charging longitudinal interval CZ j , when the number of interval charging peak points in all curves within the charging longitudinal interval CZ j is greater than 1, mark the interval charging peak point with the largest slope among all interval charging peak points as the preferred charging peak point; when the number of interval charging peak points in all curves within the charging longitudinal interval CZ j is equal to 1, mark the interval charging peak point within the charging longitudinal interval CZ j as the preferred charging peak point; when the number of interval charging peak points in all curves within the charging longitudinal interval CZ j is equal to 0, mark the point with the largest slope among all curves within the charging longitudinal interval CZ j as the preferred charging peak point; Obtain the preferred charging peak points corresponding to all charging longitudinal intervals CZ; Denote all charging longitudinal intervals CZ and their preferred charging peak points as the energy storage unit CN z of the charging optimization parameters.

5. The energy storage configuration optimization method based on deep learning according to claim 4, wherein Establish a discharge test based on the way of obtaining the charging optimization parameters, and conduct a discharge test on the energy storage device based on all discharge rates capable of discharging the energy storage unit. Obtaining the discharge optimization parameters of each energy storage unit in the energy storage device based on the results of the discharge test includes: Perform a discharge test on the energy storage unit CN z The discharge test is as follows: Based on the charge test, obtain all the discharge rates at which the energy storage unit CN can be discharged. After fully charging the electrical energy in the energy storage unit CN, sequentially discharge the charging unit CN using all the discharge rates, and establish a rectangular coordinate system to obtain the curves corresponding to all the discharge rates; Analyze the curves corresponding to all discharge rates based on the acquisition method of the charging optimization parameters, denote the interval obtained by the straight line parallel to the X-axis as the discharge longitudinal interval FZ, and denote the point obtained based on the preferred charging peak point as the preferred discharge peak point; denote all the discharge longitudinal intervals FZ and their preferred discharge peak points as the energy storage unit CN z The discharge optimization parameters; Obtain the charging optimization parameters and the discharge optimization parameters of all energy storage units CN.

6. The method for optimizing energy storage configuration based on deep learning according to claim 5, wherein Conducting a primary optimization of the configuration of the energy storage device based on the charging optimization parameters and the discharge optimization parameters includes: The initial optimization includes: when charging the energy storage battery, obtaining the electric energy of each energy storage unit CN in the energy storage battery in real time; for any unfilled energy storage unit CN z , obtaining the charging optimization parameters of the charging unit CN z , and recording the charging longitudinal interval CZ where the electric energy in the energy storage unit CN z is located as the preferred charging interval, and recording the chargeable rate CS corresponding to the curve where the preferred charging peak point of the energy storage unit CN z in the preferred charging interval is located as the preferred charging rate, and using the preferred charging rate to charge the energy storage unit CN z . When discharging the energy storage battery, the electric energy of each energy storage unit CN in the energy storage battery is obtained in real time; for any unfilled energy storage unit CN z , the discharge optimization parameters of the charging unit CN z are obtained, and the discharge longitudinal interval FZ where the electric energy in the energy storage unit CN z is located is recorded as the preferred discharge interval, and the discharge rate corresponding to the curve where the preferred discharge peak point of the energy storage unit CN z in the preferred discharge interval is located is recorded as the preferred discharge rate, and the energy storage unit CN z is discharged using the preferred discharge rate.

7. The method for optimizing energy storage configuration based on deep learning according to claim 6, wherein Establish a deep learning model, train the deep learning model based on the acquisition of the charging optimization parameters and the discharge optimization parameters, and using the trained deep learning model to conduct real-time optimization of the energy storage device includes: Establish a deep learning model, store the charge and discharge data of the energy storage battery after primary optimization in the deep learning model. The unit at the input end of the deep learning model is used to input the type of electricity consumption and the remaining electric energy, and the unit at the output end of the deep learning model is used to output the charge and discharge rate. Among them, the type of electricity consumption includes charging and discharging, and the charge and discharge rate includes the chargeable rate and the discharge rate; Conduct real-time optimization of the charging optimization parameters and the discharge optimization parameters of each energy storage unit CN based on the charge and discharge data of the energy storage battery stored in the deep learning model. The optimization method is: When the energy storage unit CN is charging or discharging, input the type of electricity consumption and the remaining electric energy of the energy storage unit CN into the deep learning model. Based on the simulation of all chargeable rates or discharge rates, obtain the fastest chargeable rate or discharge rate when the energy storage unit CN is fully charged or emptied from the remaining electric energy, and output it by the unit at the output end; When the output result of the deep learning model is the same as the preferred charging rate or the preferred discharge rate obtained by using primary optimization, do not optimize the charging optimization parameters and the discharge optimization parameters of the energy storage unit CN; When the output result of the deep learning model is different from the preferred charging rate or the preferred discharging rate obtained by the initial optimization, an optimization selection algorithm is used to obtain an optimization judgment value. The optimization selection algorithm is as follows: , where G is the optimization judgment value, f0 is the output result of the deep learning model. When the output result of the deep learning is the chargeable rate, g is the number of preferred charging rates obtained by the initial optimization for the energy storage unit CN, and f i is the i-th preferred charging rate obtained by the initial optimization for the energy storage unit CN; when the output result of the deep learning is the discharging rate, g is the number of preferred discharging rates obtained by the initial optimization for the energy storage unit CN, and f i is the i-th preferred discharging rate obtained by the initial optimization for the energy storage unit CN; When the optimization judgment value is greater than 0, the output result of the deep learning model is used to replace the preferred charging rate or the preferred discharging rate of the energy storage unit CN; When the optimization judgment value is less than or equal to 0, the charging optimization parameters and the discharging optimization parameters of the energy storage unit CN are not optimized.

8. A deep learning-based energy storage configuration optimization system for implementing the deep learning-based energy storage configuration optimization method according to any one of claims 1-7, characterized in that, It includes a charging optimization analysis module, a discharging optimization analysis module, and a deep optimization module; The charging optimization analysis module is used to obtain multiple charging rates capable of charging the energy storage device, and perform a charging test on the energy storage device based on the multiple charging rates; Based on the results of the charging test, obtain the charging optimization parameters of each energy storage unit in the energy storage device; The discharging optimization analysis module is used to establish a discharging test based on the obtaining method of the charging optimization parameters, and perform a discharging test on the energy storage device based on all discharging rates capable of discharging the energy storage unit. Based on the results of the discharging test, obtain the discharging optimization parameters of each energy storage unit in the energy storage device; Based on the charging optimization parameters and the discharging optimization parameters, perform an initial optimization on the configuration of the energy storage device; The deep optimization module is used to establish a deep learning model, and train the deep learning model based on the obtaining of the charging optimization parameters and the discharging optimization parameters, and use the trained deep training model to perform real-time optimization on the energy storage device.

9. An electronic device, characterized in that, It includes a processor and a memory. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the steps in the method according to any one of claims 1-7 are run.

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

Images