A SOC-balancing-based distributed energy storage control method and system

Abstract

The application relates to the technical field of energy storage control, and particularly discloses a distributed energy storage control method and system based on SOC balancing, which comprises the following steps: collecting real-time energy storage state data and real-time environmental data, determining the real-time state feature vector of each energy storage unit and the real-time environmental feature vector; obtaining and analyzing historical operation data of the distributed energy storage system, determining the execution field set of each energy storage unit and a plurality of execution control cluster sets of each historical execution field in the execution field set; determining the multiplexing instruction set and the correction instruction set of each energy storage unit; and determining the real-time instruction set of each energy storage unit. The adaptability of the instruction to the current working condition can be improved, the instruction deviation can be reduced, the control comprehensiveness and accuracy can be enhanced, the SOC balancing speed and precision can be improved, the adaptability to complex environments can be enhanced, energy waste can be reduced, the utilization rate of energy storage resources can be improved, the stable operation of the system can be ensured, the service life of equipment can be prolonged, and the operation and maintenance cost can be reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy storage control, in particular to a distributed energy storage control method and system based on SOC balancing. BACKGROUND

[0002] Early distributed energy storage control relies on simple logic and does not pay attention to SOC balancing, resulting in large differences in the state of each energy storage unit, low overall system efficiency, and short service life. With the expansion of distributed energy storage scale, centralized control SOC balancing methods have emerged, which can manage uniformly but have heavy communication burden and slow response. Once the main controller fails, the system operation is prone to interruption.

[0003] After that, distributed control emerged, each unit can adjust autonomously, communication cost is reduced, and system efficiency is improved. However, this method relies on pre-set rules and lacks deep mining and utilization of historical data, making it difficult to adapt to complex and variable working conditions. In the face of environmental changes and load fluctuations, it is difficult to adjust the control strategy flexibly, and the SOC balancing effect is not good, which cannot fully play the performance of the distributed energy storage system.

[0004] Therefore, the present application provides a distributed energy storage control method and system based on SOC balancing. SUMMARY

[0005] The present application provides a distributed energy storage control method and system based on SOC balancing, determines the real-time state feature vector and real-time environment feature vector of each energy storage unit according to the collected real-time energy storage state data and real-time environment data, determines the execution field set of each energy storage unit and the multiple execution control cluster sets of each historical execution field in the execution field set by analyzing the obtained historical operation data, and determines the reuse instruction set, the correction instruction set and the real-time instruction set of each energy storage unit, and executes the real-time instruction set of all energy storage units. It can accurately adapt to the working conditions of each energy storage unit. It can improve the adaptability of the instruction to the current working condition, reduce the instruction deviation, enhance the comprehensiveness and accuracy of the control, improve the SOC balancing speed and accuracy, enhance the adaptability to complex environment, reduce energy waste, improve the utilization rate of energy storage resources, ensure stable operation of the system, prolong the service life of the equipment, and reduce the operation and maintenance cost.

[0006] The present application provides a distributed energy storage control method based on SOC balancing, comprising:

[0007] S1: Collecting real-time energy storage state data and real-time environment data of all energy storage units in the distributed energy storage system, and determining the real-time state feature vector and real-time environment feature vector of each energy storage unit;

[0008] S2: obtaining and analyzing historical operation data of the distributed energy storage system, determining an execution field set of each energy storage unit and a plurality of execution control cluster sets of each historical execution field in the execution field set;

[0009] S3: determining a multiplexing instruction set and a correction instruction set of each energy storage unit based on the execution control cluster set and all execution control cluster sets of each historical execution field in the execution field set of each energy storage unit;

[0010] S4: determining a real-time instruction set of each energy storage unit based on the multiplexing instruction set and the correction instruction set of each energy storage unit, and implementing distributed energy storage control based on the real-time instruction sets of all energy storage units.

[0011] Preferably, a distributed energy storage control method based on SOC balancing, collects real-time energy storage state data and real-time environment data of all energy storage units in the distributed energy storage system, including:

[0012] Collecting real-time energy storage state sub-data of each energy storage unit in the current control period based on the state sensor group installed on each energy storage unit;

[0013] Preprocessing the real-time energy storage state sub-data of each energy storage unit, and determining the real-time energy storage state data based on all preprocessed real-time energy storage state sub-data;

[0014] Collecting real-time environment sub-data of each energy storage unit in the current control period based on the environment sensor group installed on each energy storage unit, wherein the real-time environment sub-data includes a plurality of environment parameters and real-time environment values of each environment parameter;

[0015] Preprocessing the real-time environment sub-data of each energy storage unit, and determining the real-time environment data based on all preprocessed real-time environment sub-data.

[0016] Preferably, a distributed energy storage control method based on SOC balancing, determines a real-time state feature vector and a real-time environment feature vector of each energy storage unit, including:

[0017] Feature extraction is performed on the real-time energy storage state sub-data of each energy storage unit in the real-time energy storage state data to determine the real-time state feature vector of each energy storage unit;

[0018] Determining the real-time environment feature vector of the energy storage unit based on all real-time environment values in the real-time environment sub-data of each energy storage unit in the real-time environment data.

[0019] Preferably, a distributed energy storage control method based on SOC balancing, obtains and analyzes historical operation data of a distributed energy storage system, determines an execution field set of each energy storage unit and a plurality of execution control cluster sets of each historical execution field in the execution field set, including:

[0020] Determine a historical specified period based on a current date corresponding to a current control period and a set historical quantity;

[0021] Obtain historical operation sub-data of the distributed energy storage system based on each historical control period in the historical specified period, wherein the historical operation sub-data includes a historical state feature vector, a historical environment feature vector of each energy storage unit, and historical instruction sub-data based on SOC balancing, and the historical instruction sub-data includes a plurality of historical execution instructions, a historical execution field and an execution value of each historical execution instruction;

[0022] Determine the historical operation data of the distributed energy storage system based on the historical operation sub-data of all historical control periods in the historical specified period.

[0023] Preferably, a distributed energy storage control method based on SOC balancing, obtains and analyzes historical operation data of a distributed energy storage system, determines an execution field set of each energy storage unit and a plurality of execution control cluster sets of each historical execution field in the execution field set, further including:

[0024] Determine the execution field set of each energy storage unit based on the historical execution field of all historical execution instructions of each energy storage unit in the historical instruction sub-data of all historical operation sub-data in the historical operation data, and determine the execution frequency and the execution control period set of each historical execution field in the execution field set of each energy storage unit, wherein the execution control period set includes a plurality of historical control periods;

[0025] Cluster analyze the execution control period set of each historical execution field in the execution field set of each energy storage unit based on the historical state feature vector and the historical environment feature vector of all historical control periods in the execution control period set of each historical execution field in the execution field set of each energy storage unit, and determine a plurality of execution control cluster sets of each historical execution field in the execution field set of each energy storage unit based on the cluster analysis result, wherein the execution control cluster set includes a plurality of historical control periods.

[0026] Preferably, a distributed energy storage control method based on SOC balancing, determines a multiplexing instruction set and a correction instruction set of each energy storage unit based on the execution control cluster set and all execution control cluster sets of each historical execution field in the execution field set of each energy storage unit, including:

[0027] performing polynomial fitting on each state feature value of each state feature vector of all historical control periods in the execution control cluster set of each historical execution field in the execution field set of each energy storage unit, to determine a real-time state fitting value of each state feature of the historical state feature vector based on each execution control cluster set of each energy storage unit;

[0028] performing polynomial fitting on each environment feature value of each environment feature vector of all historical control periods in the execution control cluster set of each historical execution field in the execution field set of each energy storage unit, to determine a real-time environment fitting value of each environment feature of the historical environment feature vector based on each execution control cluster set of each energy storage unit;

[0029] performing polynomial fitting on each execution value of the historical execution instruction of the historical execution field of all historical control periods in the execution control cluster set of each historical execution field in the execution field set of each energy storage unit, to determine an execution fitting value of each historical execution field based on each execution control cluster set of each energy storage unit;

[0030] based on the historical running data, the execution field set of each energy storage unit, the execution frequency of each historical execution field in the execution field set of each energy storage unit, the real-time state fitting value of all state features of the historical state feature vector based on all execution control cluster sets of each energy storage unit, and the real-time environment fitting value of all environment features of the historical environment feature vector, calculating the instruction label and the cluster label of each historical execution field in the execution field set of each energy storage unit, wherein the instruction label includes a reuse label and a correction label;

[0031] based on all historical execution fields in the execution field set of each energy storage unit with a reuse label, determining a reuse instruction set of each energy storage unit;

[0032] based on all historical execution fields in the execution field set of each energy storage unit with a correction label, determining a correction instruction set of each energy storage unit.

[0033] Preferably, a distributed energy storage control method based on SOC balancing, based on the reuse instruction set and the correction instruction set of each energy storage unit, determining a real-time instruction set of each energy storage unit, and based on the real-time instruction set of all energy storage units, implementing distributed energy storage control, including:

[0034] based on each historical execution field in the reuse instruction set of each energy storage unit and the execution fitting value of each historical execution field, determining a real-time execution instruction of each historical execution field in the reuse instruction set of each energy storage unit;

[0035] based on each historical execution field in the modified instruction set of each energy storage unit and the execution adjustment value of each historical execution field, an execution adjustment value of each historical execution field in the modified instruction set of each energy storage unit is calculated;

[0036] based on each historical execution field in the modified instruction set of each energy storage unit and the execution adjustment value of each historical execution field, an execution adjustment value of each historical execution field in the modified instruction set of each energy storage unit is calculated;

[0037] based on the real-time energy storage state sub-data, the real-time environment data, the real-time execution instruction of all historical execution fields in the multiplexed instruction set of each energy storage unit, and the real-time execution instruction of each historical execution field in the multiplexed instruction set of each energy storage unit, a plurality of supplementary execution instructions of each energy storage unit are generated;

[0038] based on the real-time execution instruction of all historical execution fields in the multiplexed instruction set of each energy storage unit, the real-time execution instruction of all historical execution fields in the multiplexed instruction set, and all supplementary execution instructions, a real-time instruction set of each energy storage unit is determined;

[0039] the real-time instruction set of all energy storage units is executed to realize distributed energy storage control of the distributed energy storage system.

[0040] The application provides a distributed energy storage control system based on SOC balancing, which is used to execute any one of the distributed energy storage control methods based on SOC balancing in embodiments 1 to 7, and comprises:

[0041] The acquisition module acquires real-time energy storage state data and real-time environment data of all energy storage units in the distributed energy storage system, and determines real-time state feature vectors of each energy storage unit and real-time environment feature vectors.

[0042] The analysis module acquires and analyzes historical operation data of the distributed energy storage system, and determines an execution field set of each energy storage unit and a plurality of execution control cluster sets of each historical execution field in the execution field set.

[0043] The instruction module determines a multiplexed instruction set and a modified instruction set of each energy storage unit based on the execution control cluster sets and all execution control cluster sets of each historical execution field in the execution field set of each energy storage unit.

[0044] The control module determines a real-time instruction set of each energy storage unit based on the multiplexed instruction set and the modified instruction set of each energy storage unit, and realizes distributed energy storage control based on the real-time instruction set of all energy storage units.

[0045] The beneficial effects generated by the present application relative to the prior art are: according to the collected real-time energy storage state data and real-time environment data, the real-time state feature vector of each energy storage unit and the real-time environment feature vector are determined, the historical operation data obtained by analysis is used to determine the execution field set of each energy storage unit and the multiple execution control cluster sets of each historical execution field in the execution field set, and the multiplexing instruction set, the correction instruction set and the real-time instruction set of each energy storage unit are determined, and the real-time instruction set of all energy storage units is executed. The working conditions of each energy storage unit can be accurately adapted. The adaptability of the instruction to the current working condition can be improved, the instruction deviation can be reduced, the control comprehensiveness and accuracy can be enhanced, the SOC balancing speed and accuracy can be improved, the adaptability to complex environment can be enhanced, the energy waste can be reduced, the energy storage resource utilization rate can be improved, the stable operation of the system can be ensured, the equipment life can be prolonged, and the operation and maintenance cost can be reduced.

[0046] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and attained by the structure particularly pointed out in the specification.

[0047] The technical solutions of the present application will be further described in detail below with the help of the accompanying drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0048] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, together with the embodiments of the present application, to explain the present application, and do not constitute a limitation on the present application. In the drawings:

[0049] Figure 1 A SOC balancing based distributed energy storage control method is provided in an embodiment of the present application.

[0050] Figure 2 A SOC balancing based distributed energy storage control system flow chart is provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] The preferred embodiments of the present application will be described below in conjunction with the accompanying drawings, and it should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and do not limit the present application. Embodiment 1

[0052] The present application provides a SOC balancing based distributed energy storage control method, which refers to Figure 1 , comprising:

[0053] S1: Collecting real-time energy storage state data and real-time environment data of all energy storage units in the distributed energy storage system, and determining the real-time state feature vector of each energy storage unit and the real-time environment feature vector;

[0054] S2: Obtain and analyze historical operation data of the distributed energy storage system, determine an execution field set of each energy storage unit and a plurality of execution control cluster sets of each historical execution field in the execution field set;

[0055] S3: Determine a reuse instruction set and a correction instruction set of each energy storage unit based on the execution control cluster sets and all execution control cluster sets of each historical execution field in the execution field set of each energy storage unit;

[0056] S4: Determine a real-time instruction set of each energy storage unit based on the reuse instruction set and the correction instruction set of each energy storage unit, and implement distributed energy storage control based on the real-time instruction sets of all energy storage units.

[0057] In this embodiment, the current real-time energy storage state data of all energy storage units in the system, such as SOC, current, voltage, etc., and real-time environmental data, such as temperature, humidity, etc., are collected. Then, key information is extracted from these data to construct a real-time state feature vector and a real-time environmental feature vector for each energy storage unit, which can simply reflect the current state and environment of the energy storage unit.

[0058] In this embodiment, the past operation data of the distributed energy storage system is first obtained, which contains the state, environment and executed instructions of each energy storage unit at different times. By deeply analyzing these historical data, the execution fields of various types of instructions that each energy storage unit has ever executed are sorted out to form an execution field set. Then, for each historical execution field in the set, clustering is performed according to its state, environment and other features when executed, to obtain a plurality of execution control cluster sets, each of which contains a plurality of historical control periods with similar working conditions.

[0059] In this embodiment, based on the obtained execution control cluster sets, the applicability of these historical execution fields under the current and similar working conditions is analyzed in combination with all cluster sets corresponding to each historical execution field. For those historical execution fields that perform well in similar working conditions and can be directly applied, they are classified into a reuse instruction set. For those that need to be adjusted according to the actual situation to be applicable, they are classified into a correction instruction set.

[0060] In this embodiment, the reuse instruction set and the correction instruction set of each energy storage unit are integrated, and the real-time state and environmental feature vectors are combined to generate a real-time instruction set of each energy storage unit in the current control period. These real-time instruction sets cover the specific operations that each energy storage unit needs to perform. Finally, the system performs control operations according to the real-time instruction sets of all energy storage units to achieve SOC balance among the units and ensure efficient and stable operation of the entire distributed energy storage system.

[0061] The beneficial effects of the above technology are: according to the collected real-time energy storage state data and real-time environment data, the real-time state feature vector of each energy storage unit and the real-time environment feature vector are determined, the historical operation data obtained by analysis is used to determine the execution field set of each energy storage unit and the multiple execution control cluster sets of each historical execution field in the execution field set, and the reuse instruction set, the correction instruction set and the real-time instruction set of each energy storage unit are determined, and the real-time instruction set of all energy storage units is executed. It can accurately adapt to the working condition of each energy storage unit. It can improve the adaptability of the instruction to the current working condition, reduce the instruction deviation, enhance the control comprehensiveness and accuracy, improve the SOC balancing speed and accuracy, enhance the adaptability to complex environment, reduce energy waste, improve the utilization rate of energy storage resources, ensure the stable operation of the system, prolong the service life of the equipment, and reduce the operation and maintenance cost. Embodiment 2

[0062] Based on embodiment 1, a distributed energy storage control method based on SOC balancing, collecting real-time energy storage state data and real-time environment data of all energy storage units in the distributed energy storage system, comprising:

[0063] Collecting real-time energy storage state sub-data of each energy storage unit in the current control period based on the state sensor group installed on each energy storage unit;

[0064] Preprocessing the real-time energy storage state sub-data of each energy storage unit, and determining the real-time energy storage state data based on all preprocessed real-time energy storage state sub-data;

[0065] Collecting real-time environment sub-data of each energy storage unit in the current control period based on the environment sensor group installed on each energy storage unit, wherein the real-time environment sub-data includes multiple environment parameters and real-time environment values of each environment parameter;

[0066] Preprocessing the real-time environment sub-data of each energy storage unit, and determining the real-time environment data based on all preprocessed real-time environment sub-data.

[0067] In this embodiment, for each energy storage unit, the real-time energy storage state sub-data in the current control period is collected by relying on the state sensor group installed on itself. The state sensor group here is specially designed for monitoring the running state of the energy storage unit, which may include current sensors, voltage sensors, SOC (state of charge) monitoring sensors, etc. These sensors will continuously collect data related to the running state of the energy storage unit in the current control period, such as charge and discharge current value, real-time voltage value, current SOC estimation value, etc.

[0068] In this embodiment, the real-time energy storage state sub-data collected by each energy storage unit is pre-processed. The pre-processing link is mainly to eliminate interference information in the original sub-data, such as abnormal jump values of the sensor due to electromagnetic interference, missing data during collection, etc. Specifically, abnormal values that are obviously beyond the reasonable range (such as the current sensor suddenly collecting a value far exceeding the rated current of the energy storage unit, which can be determined as an abnormal value and eliminated), missing data can be supplemented by a reasonable interpolation method (such as missing SOC data at a certain time, which can be estimated and supplemented according to the SOC data trend of the adjacent time), and the data can also be smoothed to reduce the error caused by short-term fluctuations (such as averaging the current values collected multiple times in a control period to obtain more stable current data in that period).

[0069] In this embodiment, each energy storage unit is equipped with a special environment sensor group, and these environment sensor groups will collect the real-time environment sub-data of the environment where the energy storage unit is located in the current control period. The environment sensor group usually includes temperature sensors, humidity sensors, light sensors, etc. (the specific sensor type is adjusted according to the installation scene of the energy storage unit, such as wind speed sensors for outdoor energy storage units), and the collected real-time environment sub-data not only includes multiple environmental parameters (such as temperature, humidity, light intensity, etc.), but also records the real-time environment value corresponding to each environmental parameter (such as the real-time value of the temperature parameter is 25°C, the real-time value of the humidity parameter is 60%, etc.), which together constitute the real-time environment sub-data of the energy storage unit.

[0070] In this embodiment, the real-time environment sub-data of each energy storage unit is pre-processed. The purpose of this step is similar to the pre-processing of real-time energy storage state sub-data, which is to improve data quality. For example, the temperature sensor may occasionally drift due to device aging, and these abnormal values need to be corrected by comparing with other normal sensor data in the same area; the humidity sensor may have data lag in high humidity environment, and the data needs to be adjusted through time series analysis to make it more consistent with the current actual humidity; for the missing data of a certain environmental parameter caused by occasional sensor failure, it will also be supplemented by a reasonable method (such as referring to the same type of environmental parameter value of the energy storage unit in the surrounding similar position), to ensure that the real-time environment value of each environmental parameter is accurate and complete.

[0071] The beneficial effects of the above technology are: collecting the real-time energy storage state data and real-time environment data of all energy storage units in the distributed energy storage system can ensure the accuracy and integrity of the real-time data, and provide high-quality data support for subsequent control decisions. Embodiment 3:

[0072] On the basis of embodiment 2, a distributed energy storage control method based on SOC balancing, the real-time state feature vector of each energy storage unit and the real-time environment feature vector are determined, including:

[0073] The real-time energy storage state data of each energy storage unit is extracted, and the real-time state feature vector of each energy storage unit is determined;

[0074] Based on all real-time environment values in the real-time environment sub-data of each energy storage unit in the real-time environment data, the real-time environment feature vector of the energy storage unit is determined.

[0075] In this embodiment, the real-time energy storage state data contains various real-time energy storage state sub-data of each energy storage unit. From these real-time energy storage state sub-data, the key information that best reflects the core state of the energy storage unit is screened and refined. For example, for current-related sub-data, instead of simply listing all collected current values, representative features such as average current, maximum current, and current change trend are extracted; for SOC sub-data, features such as current SOC value and SOC change rate may be extracted. After such extraction process, the real-time energy storage state sub-data of each energy storage unit is converted into a vector composed of key features, that is, a real-time state feature vector.

[0076] In this embodiment, the real-time environment data of each energy storage unit contains multiple environment parameters and their corresponding real-time environment values, such as temperature value, humidity value, and light intensity value. These environment parameters and values together constitute the environment information of the energy storage unit. When determining the real-time environment feature vector, the real-time environment values of these environment parameters need to be sorted and integrated. This vector can reflect the current environment of the energy storage unit.

[0077] The beneficial effects of the above technology are: determining the real-time state feature vector of each energy storage unit and the real-time environment feature vector can highlight key features, reduce redundant information interference, and improve data processing efficiency. Embodiment 4:

[0078] On the basis of embodiment 1, a distributed energy storage control method based on SOC balancing, the historical running data of the distributed energy storage system is obtained and analyzed, the execution field set of each energy storage unit and the multiple execution control cluster sets of each historical execution field in the execution field set are determined, including:

[0079] Determine the historical specified period based on the current date corresponding to the current control period and the set historical number;

[0080] The historical operation sub-data of each historical control period in the historical specified period is acquired, wherein the historical operation sub-data includes a historical state feature vector of each energy storage unit, a historical environment feature vector, and historical instruction sub-data based on SOC balancing, the historical instruction sub-data includes a plurality of historical execution instructions, and a historical execution field and an execution value of each historical execution instruction;

[0081] The historical operation data of the distributed energy storage system is determined based on the historical operation sub-data of all historical control periods in the historical specified period.

[0082] In this embodiment, the historical specified period is determined. The core of this step is to combine two key information: the current date corresponding to the current control period and the set historical quantity. The current date specifies the time reference point, such as August 15th. The set historical quantity specifies the number of historical control periods to be traced back, such as setting the historical quantity to 30 days, which means all historical control periods within 30 days are included. Through these two information, a specific time range as the historical specified period can be determined, such as all control periods within 30 days from July 16th to August 14th. This clearly defines the time limit for acquiring historical data in the subsequent step, ensuring that the selected historical period has a certain correlation with the current period in time, facilitating meaningful analysis and comparison in the subsequent step.

[0083] In this embodiment, the historical operation sub-data is acquired. Within the determined historical specified period, the historical operation sub-data of the distributed energy storage system is collected for each historical control period. These sub-data include the historical state feature vector of each energy storage unit, which is a collection of key features reflecting the state of the energy storage unit in a past control period, such as the SOC, current, voltage, and other feature information at a certain time in the past; the historical environment feature vector, which records the key features of the environment in which the energy storage unit is located in the corresponding historical control period, such as temperature, humidity, and other environmental parameters; and the historical instruction sub-data based on SOC balancing, which details the instructions related information executed in the past to achieve SOC balancing. Specifically, it includes a plurality of historical execution instructions, each instruction having a corresponding historical execution field (which can be understood as descriptive information of the specific operation content or operation object of the instruction) and an execution value (which is the specific parameter value when the instruction is executed, such as charging and discharging power, duration, etc.).

[0084] In this embodiment, after collecting the historical operation sub-data of all historical control periods in the historical specified period, these sub-data need to be integrated and processed. The historical operation data comprehensively records the state of each energy storage unit in the system, the environment in which it is located, and the instructions executed to achieve SOC balancing, etc. in the selected historical time period.

[0085] The beneficial effects of the above technology are: obtaining and analyzing historical operation data of the distributed energy storage system, determining the execution field set of each energy storage unit and the multiple execution control cluster sets of each historical execution field in the execution field set, which can accurately circumscribe the relevant historical data range, improve data utilization efficiency, and provide more targeted historical reference for decision-making. Embodiment 5:

[0086] Based on embodiment 4, a distributed energy storage control method based on SOC balancing, obtaining and analyzing historical operation data of the distributed energy storage system, determining the execution field set of each energy storage unit and the multiple execution control cluster sets of each historical execution field in the execution field set, further comprising:

[0087] Based on the historical execution field of all historical execution instructions of each energy storage unit in the historical instruction sub-data in all historical running sub-data in the historical operation data, the execution field set of each energy storage unit is determined, and the execution frequency and the execution control period set of each historical execution field in the execution field set of each energy storage unit are determined, wherein the execution control period set includes multiple historical control periods;

[0088] Based on the historical state feature vector and the historical environment feature vector of all historical control periods in the execution control period set of each historical execution field in the execution field set of each energy storage unit, the execution control period set of each historical execution field in the execution field set of each energy storage unit is analyzed, and based on the clustering analysis result, multiple execution control cluster sets of each historical execution field in the execution field set of each energy storage unit are determined, wherein the execution control cluster set includes multiple historical control periods.

[0089] In this embodiment, starting from the historical instruction sub-data in all historical running sub-data in the historical operation data, the historical execution field of all historical execution instructions of each energy storage unit is focused. These historical execution fields are the parts in each historical execution instruction that describe the specific operation content or object, such as constant current charging, "equalization discharge, etc. Collecting all different historical execution fields belonging to the same energy storage unit forms the execution field set of the energy storage unit, which covers the operation description of various instructions that the energy storage unit has ever executed.

[0090] In this embodiment, for each historical execution field in the execution field set of each energy storage unit, the execution frequency and the execution control period set thereof are determined. The execution frequency refers to the proportion of the number of times that the historical execution field is executed in all historical control periods, for example, if a certain execution field appears 80 times in 100 historical control periods, the execution frequency thereof is high. The execution control period set refers to the collection of all historical control periods in which the historical execution field is executed, for example, the execution field is executed in the 5th, 10th and 15th historical control periods, and these periods collectively constitute the execution control period set thereof.

[0091] In this embodiment, for each historical execution field in the execution field set of each energy storage unit, the execution control period set thereof is used to determine the historical state feature vector and the historical environment feature vector corresponding to all historical control periods. That is, these historical control periods are grouped according to the similarity of their historical state feature vectors and historical environment feature vectors. Historical control periods with high similarity, that is, similar state and environment conditions, are grouped together. Based on the clustering analysis results, the original execution control period set is divided into multiple sub-sets, and each sub-set is an execution control cluster set. The multiple historical control periods in each execution control cluster set have similar historical state and historical environment features, which means that the corresponding historical execution field is executed under similar working conditions.

[0092] The above-mentioned technology has the following beneficial effects: obtaining and analyzing historical operation data of a distributed energy storage system, determining the execution field set of each energy storage unit and multiple execution control cluster sets of each historical execution field in the execution field set, accurately mining the execution rules of historical instructions, reducing invalid data interference, highlighting the internal relationship between instructions and working conditions, improving the utilization efficiency of historical data, providing a structured basis for subsequent instruction reuse and correction, and enhancing the judgment ability of instruction applicability under different working conditions. Embodiment 6

[0093] Based on embodiment 5, a distributed energy storage control method based on SOC balancing determines the reuse instruction set and the correction instruction set of each energy storage unit based on the execution control cluster set and all execution control cluster sets of each historical execution field in the execution field set of each energy storage unit, including:

[0094] For each historical execution field in the execution field set of each energy storage unit, the feature values of each state feature of the historical state feature vector of all historical control periods in the execution control cluster set thereof are subjected to quadratic polynomial fitting to determine the real-time state fitting value of each state feature of the historical state feature vector for each execution control cluster set of each energy storage unit.

[0095] The characteristic values of each environment feature of the historical environment feature vectors of all historical control periods in the execution control cluster set of each historical execution field of the execution field set of each energy storage unit are subjected to quadratic polynomial fitting to determine the real-time environment fitting values of each execution control cluster set of each energy storage unit based on each environment feature of the historical environment feature vectors;

[0096] The execution values of the historical execution instructions of the historical execution field of all historical control periods in the execution control cluster set of each historical execution field of the execution field set of each energy storage unit are subjected to quadratic polynomial fitting to determine the execution fitting values of each execution control cluster set of each energy storage unit based on each historical execution field;

[0097] The instruction tags and cluster tags of each historical execution field in the execution field set of each energy storage unit are calculated based on the historical running data, the execution frequency of each historical execution field in the execution field set of each energy storage unit, the real-time state fitting values of all state features of the historical state feature vectors and the real-time environment fitting values of all environment features of the historical environment feature vectors of each execution control cluster set of each energy storage unit, wherein the instruction tags include multiplexing tags and correction tags;

[0098] The multiplexing instruction set of each energy storage unit is determined based on all historical execution fields with multiplexing tags in the execution field set of each energy storage unit;

[0099] The correction instruction set of each energy storage unit is determined based on all historical execution fields with correction tags in the execution field set of each energy storage unit.

[0100] In this embodiment, each execution control cluster set of each historical execution field in the execution field set of each energy storage unit contains multiple historical control periods, each of which has a historical state feature vector composed of multiple state features and corresponding characteristic values. The characteristic values of each state feature are subjected to quadratic polynomial fitting, for example, weighted or regularized least squares, moving window quadratic fitting, etc., to obtain a quadratic polynomial that best reflects the variation law of these characteristic values. The result obtained through this polynomial is the real-time state fitting value of each execution control cluster set based on this state feature, and each state feature has a corresponding fitting value.

[0101] In this embodiment, similarly, in each execution control cluster set, the historical environment feature vector of each historical control period contains multiple environment features and characteristic values. The characteristic values of each environment feature are subjected to quadratic polynomial fitting, and the result obtained is the real-time environment fitting value of each execution control cluster set based on this environment feature, and each environment feature corresponds to a fitting value.

[0102] In this embodiment, each execution control cluster set has execution values of all historical execution instructions corresponding to the historical execution fields of all historical control periods. The execution values are fitted by a second-degree polynomial to obtain the execution fitting values of the execution control cluster set based on the historical execution field.

[0103] In this embodiment, based on the historical running data, the execution field set of each energy storage unit, the execution frequency of each historical execution field in the execution field set of each energy storage unit, the real-time state fitting value of all state features based on the historical state feature vector and the real-time environment fitting value of all environment features based on the historical environment feature vector of all execution control cluster sets of each energy storage unit, the instruction label and the cluster label of each historical execution field in the execution field set of each energy storage unit are calculated. The calculation formula of the instruction label and the cluster label can be expressed as:

[0104] ;

[0105] Wherein, represents the instruction label of the jth historical execution field in the execution field set of the ith energy storage unit, represents the execution frequency of the jth historical execution field in the execution field set of the ith energy storage unit, TF1 represents the first frequency threshold, TF2 represents the second frequency threshold, and TS represents the fitting difference threshold, represents the cluster label of the jth historical execution field in the execution field set of the ith energy storage unit, represents the minimum fitting difference value of the jth historical execution field in the execution field set of the ith energy storage unit, represents the first correlation factor of the jth historical execution field in the execution field set of the ith energy storage unit and the a th state feature in the real-time state feature vector, represents the fitting degree of the jth historical execution field in the execution field set of the ith energy storage unit based on the cluster label, represents the anti-zero parameter, represents the state weight of the jth historical execution field in the execution field set of the ith energy storage unit based on the a th state feature in the real-time state feature vector, represents the environment weight of the jth historical execution field in the execution field set of the ith energy storage unit based on the b th environment feature in the real-time environment feature vector, represents the second correlation factor of the jth historical execution field in the execution field set of the ith energy storage unit and the b th environment feature in the real-time environment feature vector, represents the feature value of the a th state feature in the real-time state feature vector of the ith energy storage unit, represents the real-time state fitting value of the a-th state feature of the historical state feature vector based on the k-th execution control clustering set of the i-th energy storage unit, and N1 represents the number of state features in the real-time state feature vector, represents the feature value of the b-th environmental feature in the real-time environmental feature vector of the i-th energy storage unit, represents the real-time environmental fitting value of the b-th environmental feature of the historical environmental feature vector based on the k-th execution control clustering set of the i-th energy storage unit, N2 represents the number of environmental features in the real-time environmental feature vector, and iN3 represents the number of execution control clustering sets of the i-th energy storage unit.

[0106] In this embodiment, represents the real-time state fitting value of the a-th state feature of the historical state feature vector based on the k-th execution control clustering set of the i-th energy storage unit, and N1 represents the number of state features in the real-time state feature vector,

[0107] In this embodiment, the fitting degree of the j-th historical execution field in the execution field set of the i-th energy storage unit based on the clustering label The fitting function determined by performing quadratic polynomial fitting can be used to calculate the j-th historical execution field historical fitting value of each historical control period in the execution control clustering set corresponding to the clustering label of the j-th historical execution field in the execution field set of the i-th energy storage unit.

[0108] In this embodiment, the first frequency threshold TF1 can be 90%, and the second frequency threshold TF1 can be 75%.

[0109] In this embodiment, in the execution field set of each energy storage unit, all historical execution fields with multiplexing labels form a multiplexing instruction set of the energy storage unit, and all historical execution fields with correction labels form a correction instruction set of the energy storage unit.

[0110] The beneficial effects of the above technology are: based on the execution control cluster set and all execution control cluster sets of each historical execution field in the execution field set of each energy storage unit, the multiplex instruction set and the correction instruction set of each energy storage unit are determined, which can accurately capture the data change trend, improve the fitting value reliability, enhance the historical data conversion efficiency to instructions, highlight the applicability difference of instructions under different working conditions, improve the pertinence and reliability of instruction generation, and reduce invalid instruction interference. Embodiment 7:

[0111] Based on embodiment 6, a distributed energy storage control method based on SOC balance, based on the multiplex instruction set and the correction instruction set of each energy storage unit, the real-time instruction set of each energy storage unit is determined, and the distributed energy storage control is realized based on the real-time instruction set of all energy storage units, including:

[0112] Based on each historical execution field in the multiplex instruction set of each energy storage unit and the execution fitting value of each historical execution field, the real-time execution instruction of each historical execution field in the multiplex instruction set of each energy storage unit is determined;

[0113] Based on each historical execution field in the correction instruction set of each energy storage unit and the execution fitting value of each historical execution field, the execution adjustment value of each historical execution field in the correction instruction set of each energy storage unit is calculated;

[0114] Based on each historical execution field in the correction instruction set of each energy storage unit and the execution adjustment value of each historical execution field, the real-time execution instruction of each historical execution field in the multiplex instruction set of each energy storage unit is determined;

[0115] Based on the real-time energy storage state sub-data, the real-time environment data, the real-time execution instruction of all historical execution fields in the multiplex instruction set of each energy storage unit, and the real-time execution instruction of each historical execution field in the multiplex instruction set of each energy storage unit, a plurality of supplementary execution instructions of each energy storage unit are generated;

[0116] Based on the real-time execution instruction of all historical execution fields in the multiplex instruction set of each energy storage unit, the real-time execution instruction of all historical execution fields in the multiplex instruction set, and all supplementary execution instructions, the real-time instruction set of each energy storage unit is determined;

[0117] Execute the real-time instruction set of all energy storage units to realize distributed energy storage control of the distributed energy storage system.

[0118] In this embodiment, for each energy storage unit, the multiplexing instruction set contains multiple historical execution fields, and each historical execution field has an execution fitting value obtained by previous quadratic polynomial fitting. The determination of real-time execution instructions is to determine the specific instruction content to be executed in the current control period based on these historical execution fields as the basic framework and combined with the corresponding execution fitting value. For example, if a historical execution field is constant current charging and its execution fitting value is 5A, then the real-time execution instruction corresponding to this historical execution field in the current control period is to perform constant current charging at a current of 5A. It directly follows the operation type of the historical execution field and uses the execution fitting value as the key parameter of the current execution, ensuring that the instruction can be executed stably based on the historical law under the current working condition.

[0119] In this embodiment, based on each historical execution field in the correction instruction set of each energy storage unit and the execution fitting value of each historical execution field, the execution adjustment value of each historical execution field in the correction instruction set of each energy storage unit is calculated. The calculation formula of the execution adjustment value can be expressed as:

[0120] ;

[0121] Among them, represents the execution fitting value of the mth historical execution field in the correction instruction set of the ith energy storage unit, represents the jth historical execution field in the execution field set of the ith energy storage unit, represents the mth historical execution field in the correction instruction set of the ith energy storage unit, represents the value when the jth historical execution field in the execution field set of the ith energy storage unit is equal to the mth historical execution field in the correction instruction set of the ith energy storage unit, , , respectively represent the lower limit and upper limit of the execution safety of the mth historical execution field in the correction instruction set of the ith energy storage unit, represents the execution sub-adjustment value of the mth historical execution field in the correction instruction set of the ith energy storage unit, represents the execution adjustment value of the mth historical execution field in the correction instruction set of the ith energy storage unit.

[0122] In this embodiment, after obtaining the execution adjustment value of each historical execution field in the modified instruction set, the real-time execution instruction can be determined by combining the original historical execution field and the execution adjustment value. For example, in the historical execution field of constant current discharge mentioned above, the execution adjustment value is 3A, so the real-time execution instruction is to discharge at a constant current of 3A; if the historical execution field of temperature protection has an execution adjustment value of 40°C trigger power limiting, the real-time execution instruction is 40°C trigger power limiting, which adjusts the instruction to better meet the safety requirements of the current working condition.

[0123] In this embodiment, the generation of the supplementary execution instruction needs to comprehensively refer to three core bases: one is the real-time energy storage state sub-data (such as the SOC deviation of the current energy storage unit, the charge and discharge current fluctuation, etc.), the second is the real-time environmental data (such as the abnormal change of the current environmental temperature and humidity), and the third is the real-time execution instruction of the determined reuse instruction and the modified instruction (clearing the control requirements covered by the existing instructions). The core purpose is to fill the control gap not covered by the existing instructions, such as if the reuse and modified instructions only involve charge and discharge power control, and the real-time data shows that the voltage of a certain energy storage unit is close to the overvoltage threshold, the execution instruction of voltage monitoring and overvoltage shutdown is supplemented; if the real-time environmental data shows a sudden strong electromagnetic interference, and the existing instructions do not involve communication protection, the execution instruction of local standby during communication interruption is supplemented; if there is a SOC equalization gap in multiple energy storage units, and the existing instructions only target a single unit, the execution instruction of multi-unit cooperative equalization is supplemented to ensure that all potential control requirements are covered.

[0124] In this embodiment, the real-time instruction set is a summary of all instructions to be executed by the energy storage unit in the current control period, and its composition includes three parts: one is the real-time execution instruction corresponding to all historical execution fields in the reuse instruction set (directly using and adapting the basic instruction), the second is the real-time execution instruction corresponding to all historical execution fields in the modified instruction set (the optimized instruction after adjustment and adaptation), and the third is all the supplementary execution instructions determined in the above steps (guarantee instructions to fill the gap). In the summary process, these instructions will be prioritized (such as the priority of safety class supplementary instructions is higher than that of charge and discharge control instructions), and possible instruction conflicts will be excluded (such as when two instructions contradict each other on the same parameter, the instruction that better meets the safety and system goal is used), and finally a complete, conflict-free, and clear priority real-time instruction set is formed, ensuring that the energy storage unit can execute each control operation in sequence.

[0125] In this embodiment, after determining the real-time instruction set of all energy storage units, the instructions are respectively issued to the corresponding energy storage units. Each energy storage unit executes the corresponding operation according to its own real-time instruction set, such as some units performing 5A constant current charging, some units performing 3A constant current discharging, some units performing 40°C trigger limited power, and some units performing communication interruption standby. Through the synchronous execution of the real-time instructions of all energy storage units, the cooperative control of the entire distributed energy storage system is realized: on the one hand, the running state of each unit meets the safety and efficiency requirements, and on the other hand, through the cooperation of the instructions of each unit (such as discharging of high SOC units and charging of low SOC units), the SOC balancing goal at the system level is achieved, and various challenges brought by real-time environment and state changes are coped with, thereby ensuring the stable and efficient operation of the distributed energy storage system.

[0126] The beneficial effects of the above technology are that the real-time instruction set of each energy storage unit is determined based on the multiplexing instruction set and the correction instruction set of each energy storage unit, and the distributed energy storage control is realized based on the real-time instruction set of all energy storage units, which can improve the adaptability of the instructions to the current working conditions, reduce the instruction deviation, enhance the comprehensive control, realize the precise control of the system, improve the SOC balancing efficiency and the system operation stability, and reduce the energy loss. Embodiment 8:

[0127] The application provides a distributed energy storage control system based on SOC balancing, which is used for executing any one of the distributed energy storage control methods based on SOC balancing in embodiments 1 to 7, and refers to Figure 2 , and comprises:

[0128] The acquisition module acquires the real-time energy storage state data and real-time environment data of all energy storage units in the distributed energy storage system, and determines the real-time state feature vector of each energy storage unit and the real-time environment feature vector.

[0129] The analysis module acquires and analyzes the historical operation data of the distributed energy storage system, determines the execution field set of each energy storage unit and a plurality of execution control cluster sets of each historical execution field in the execution field set.

[0130] The instruction module determines the multiplexing instruction set and the correction instruction set of each energy storage unit based on the execution control cluster set and all execution control cluster sets of each historical execution field in the execution field set of each energy storage unit.

[0131] The control module determines the real-time instruction set of each energy storage unit based on the multiplexing instruction set and the correction instruction set of each energy storage unit, and realizes the distributed energy storage control based on the real-time instruction set of all energy storage units.

[0132] The beneficial effects of the above technology are: according to the collected real-time energy storage state data and real-time environment data, the real-time state feature vector of each energy storage unit and the real-time environment feature vector are determined, the historical operation data obtained is analyzed to determine the execution field set of each energy storage unit and the multiple execution control cluster sets of each historical execution field in the execution field set, and the reuse instruction set, the correction instruction set and the real-time instruction set of each energy storage unit are determined, and the real-time instruction set of all energy storage units is executed. The working conditions of each energy storage unit can be accurately adapted. The adaptability of the instruction to the current working condition can be improved, the instruction deviation can be reduced, the control comprehensiveness and accuracy can be enhanced, the SOC balancing speed and accuracy can be improved, the adaptability to complex environment can be enhanced, the energy waste can be reduced, the energy storage resource utilization rate can be improved, the system stable operation can be ensured, the equipment life can be prolonged, and the operation and maintenance cost can be reduced.

[0133] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. A distributed energy storage control method based on SOC equalization, characterized in that, include: S1: Collect real-time energy storage status data and real-time environmental data of all energy storage units in the distributed energy storage system, and determine the real-time status feature vector and real-time environmental feature vector of each energy storage unit. S2: Acquire and analyze the historical operation data of the distributed energy storage system, determine the set of execution fields for each energy storage unit, and the multiple execution control cluster sets for each historical execution field in the set of execution fields; S3: Based on the execution control cluster set and all execution control cluster sets of each historical execution field in the execution field set of each energy storage unit, determine the multiplexing instruction set and the correction instruction set of each energy storage unit; S4: Determine the real-time instruction set for each energy storage unit based on the multiplexing instruction set and the correction instruction set for each energy storage unit, and realize distributed energy storage control based on the real-time instruction set of all energy storage units; Acquire and analyze historical operational data of the distributed energy storage system to determine the set of execution fields for each energy storage unit and multiple execution control clusters for each historical execution field within the execution field set, including: The specified historical period is determined based on the current date corresponding to the current control period and the set historical quantity. Obtain historical operation sub-data of the distributed energy storage system for each historical control cycle within a specified historical period. The historical operation sub-data includes the historical state feature vector, historical environmental feature vector, and historical instruction sub-data based on SOC equalization for each energy storage unit. The historical instruction sub-data includes multiple historical execution instructions and the historical execution field and execution value of each historical execution instruction. The historical operating data of the distributed energy storage system is determined based on the historical operating sub-data of all historical control cycles within a specified historical period.

2. The distributed energy storage control method based on SOC equalization according to claim 1, characterized in that, Collect real-time energy storage status data and real-time environmental data of all energy storage units in the distributed energy storage system, including: Real-time energy storage status data of each energy storage unit is collected based on the status sensor group installed in each energy storage unit during the current control cycle; The real-time energy storage status sub-data of each energy storage unit is preprocessed, and the real-time energy storage status data is determined based on all the preprocessed real-time energy storage status sub-data. The system collects real-time environmental sub-data for each energy storage unit under the current control cycle based on the environmental sensor group installed in each energy storage unit. The real-time environmental sub-data includes multiple environmental parameters and the real-time environmental value of each environmental parameter. The real-time environmental sub-data of each energy storage unit is preprocessed, and the real-time environmental data is determined based on all the preprocessed real-time environmental sub-data.

3. The distributed energy storage control method based on SOC equalization according to claim 2, characterized in that, Determine the real-time state feature vector and real-time environmental feature vector for each energy storage unit, including: Feature extraction is performed on the real-time energy storage status sub-data of each energy storage unit in the real-time energy storage status data to determine the real-time status feature vector of each energy storage unit. Based on all real-time environmental values ​​in the real-time environmental sub-data of each energy storage unit in the real-time environmental data, the real-time environmental feature vector of the energy storage unit is determined.

4. The distributed energy storage control method based on SOC equalization according to claim 1, characterized in that, Acquire and analyze historical operational data of the distributed energy storage system, determine the set of execution fields for each energy storage unit and multiple execution control clusters for each historical execution field in the set of execution fields, and also include: Based on the historical execution fields of all historical execution instructions for each energy storage unit in the historical instruction sub-data of all historical operation data, the execution field set of each energy storage unit is determined, and the execution frequency and execution control cycle set of each historical execution field in the execution field set of each energy storage unit are determined. The execution control cycle set includes multiple historical control cycles. Based on the historical state feature vectors and historical environment feature vectors of all historical control cycles in the execution control cycle set of each historical execution field in the execution field set of each energy storage unit, cluster analysis is performed on the execution control cycle set of each historical execution field in the execution field set of each energy storage unit. Based on the cluster analysis results, multiple execution control cluster sets are determined for each historical execution field in the execution field set of each energy storage unit. The execution control cluster set includes multiple historical control cycles.

5. A distributed energy storage control method based on SOC equalization according to claim 4, characterized in that, Based on the execution control cluster set and all execution control cluster sets for each historical execution field in the execution field set of each energy storage unit, the multiplexing instruction set and correction instruction set for each energy storage unit are determined, including: For each historical execution field in the execution control cluster set of each energy storage unit, perform quadratic polynomial fitting on the feature value of each state feature of the historical state feature vector of all historical control cycles in the execution field set of each historical execution field, and determine the real-time state fitting value of each execution control cluster set of each energy storage unit based on each state feature of the historical state feature vector. For each historical execution field in the execution control cluster set of each energy storage unit, perform quadratic polynomial fitting on the feature value of each environmental feature of the historical environmental feature vector of all historical control cycles in the execution field set of each historical execution field, and determine the real-time environmental fitting value of each execution control cluster set of each energy storage unit based on each environmental feature of the historical environmental feature vector; For each historical execution field in the execution control cluster set of each historical execution field in the execution field set of each energy storage unit, perform a quadratic polynomial fitting on the execution values ​​of the historical execution instructions of the historical execution fields of all historical control cycles to determine the execution fitting value of each execution control cluster set of each energy storage unit based on each historical execution field; Based on historical operating data, the set of execution fields for each energy storage unit, the execution frequency of each historical execution field in the set of execution fields for each energy storage unit, the real-time state fitting values ​​of all state features based on historical state feature vectors for all execution control clusters of each energy storage unit, and the real-time environment fitting values ​​of all environmental features based on historical environmental feature vectors, the instruction label and cluster label of each historical execution field in the set of execution fields for each energy storage unit are calculated. The instruction label includes a reuse label and a correction label. Based on all historical execution fields in the execution field set of each energy storage unit where the instruction tag is a reuse tag, determine the set of reused instructions for each energy storage unit; Based on all historical execution fields in the execution field set of each energy storage unit that have correction tags, the correction instruction set for each energy storage unit is determined.

6. The distributed energy storage control method based on SOC equalization according to claim 5, characterized in that, The real-time instruction set for each energy storage unit is determined based on its multiplexing instruction set and correction instruction set. Distributed energy storage control is then implemented based on the real-time instruction sets of all energy storage units, including: Based on each historical execution field in the multiplexing instruction set of each energy storage unit and the execution fitting value of each historical execution field, the real-time execution instruction of each historical execution field in the multiplexing instruction set of each energy storage unit is determined. Based on each historical execution field in the correction instruction set of each energy storage unit and the execution fitting value of each historical execution field, calculate the execution adjustment value of each historical execution field in the correction instruction set of each energy storage unit; Based on each historical execution field in the correction instruction set of each energy storage unit and the execution adjustment value of each historical execution field, the real-time execution instruction for each historical execution field in the reuse instruction set of each energy storage unit is determined; Based on real-time energy storage status sub-data, real-time environmental data, real-time execution instructions for all historical execution fields in the multiplexing instruction set of each energy storage unit, and real-time execution instructions for each historical execution field in the multiplexing instruction set of each energy storage unit, multiple supplementary execution instructions for each energy storage unit are generated. Based on the real-time execution instructions of all historical execution fields in the multiplexing instruction set of each energy storage unit, the real-time execution instructions of all historical execution fields in the multiplexing instruction set, and all supplementary execution instructions, the real-time instruction set of each energy storage unit is determined. It executes a set of real-time instructions for all energy storage units to achieve distributed energy storage control of the distributed energy storage system.

7. A distributed energy storage control system based on SOC equalization, characterized in that, A method for implementing a distributed energy storage control method based on SOC equalization as described in any one of claims 1 to 6, comprising: Acquisition module: Acquires real-time energy storage status data and real-time environmental data of all energy storage units in the distributed energy storage system, and determines the real-time status feature vector and real-time environmental feature vector of each energy storage unit; Analysis module: Acquires and analyzes historical operation data of distributed energy storage system, determines the set of execution fields for each energy storage unit and multiple execution control clusters for each historical execution field in the set of execution fields; Instruction module: Based on the execution control cluster set and all execution control cluster sets of each historical execution field in the execution field set of each energy storage unit, determine the multiplexing instruction set and the correction instruction set of each energy storage unit; Control module: Based on the multiplexing instruction set and correction instruction set of each energy storage unit, the real-time instruction set of each energy storage unit is determined, and distributed energy storage control is realized based on the real-time instruction set of all energy storage units.

Images