Operation control method for flexible direct-current energy storage system in power distribution area

Abstract

The invention relates to the technical field of direct current power distribution, and provides a power distribution area flexible direct current energy storage system operation control method, which comprises the following steps of: acquiring original state evaluation data of an energy storage system in an operation state, and preprocessing the original state evaluation data to obtain state evaluation data; identifying the state evaluation data through a comprehensive evaluation model to obtain an operation state evaluation attribute of the energy storage system; identifying the state evaluation data based on a performance prediction model to obtain a performance change trend of the energy storage system; and performing dynamic control on the energy storage system through the operation state evaluation attribute and the performance change trend. According to the invention, dynamic control is realized, the adaptive capacity of the energy storage system is improved, overcharge and overdischarge are avoided, the service life of the battery is prolonged, and more accurate and more flexible operation control of the energy storage system is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of direct current power distribution, and in particular to an operation control method of a flexible direct current energy storage system in a power distribution area. Background Art

[0002] Distribution substations generally use a single-line, single-transformer power supply method, which has low reliability. There is a lack of communication between power supply in different distribution substations, and unified management is impossible. In the process of building a new power system in rural areas, a large number of distributed power sources are connected to the user side, and the traditional single power supply method cannot meet the user's electricity demand. The safe and stable operation of rural power grids faces pressures brought by problems such as decentralized access to new energy and local consumption. The power supply quality and reliability of low-voltage distribution substations, which have a direct impact on the normal production and life of residents, put forward higher requirements on the power supply stability of the substation area.

[0003] The existing flexible DC control method for substations adopts the power coordinated control of multiple energy storage clusters to achieve balanced control of the SOC of energy storage batteries. However, the balanced control of the SOC of energy storage batteries is relatively complex. The SOC of energy storage batteries needs to be maintained in a reasonable range to avoid service life degradation caused by overcharging or over-discharging. The existing technology adopts the Kalman filtering method to obtain the SOC of energy storage batteries, and then simply divides the energy storage clusters by percentage for coordinated control, ignoring other important factors that affect battery life and performance, such as temperature, charge and discharge times, and battery health status, and lacks dynamic control, resulting in poor adaptability of the energy storage system. Summary of the invention

[0004] In view of this, the present invention proposes an operation control method for a flexible direct current energy storage system in a distribution station area. Through data acquisition, preprocessing, comprehensive evaluation and performance prediction, the original state data of the energy storage system is comprehensively analyzed and processed to obtain the operation state evaluation attributes and performance change trends, thereby realizing dynamic control, improving the adaptability of the energy storage system, avoiding overcharging and over-discharging, and extending the battery life, effectively solving the problems of complex SOC balance control and lack of dynamic control in the prior art.

[0005] The technical solution of the present invention is implemented as follows: The present invention provides a method for controlling the operation of a flexible direct current energy storage system in a distribution station area, comprising the following steps:

[0006] S1, collecting original state evaluation data of the energy storage system in operation, and preprocessing the original state evaluation data to obtain state evaluation data;

[0007] S2, identifying the state assessment data through the comprehensive assessment model to obtain the operating state assessment attributes of the energy storage system;

[0008] S3, identifying the state assessment data based on the performance prediction model to obtain the performance change trend of the energy storage system;

[0009] S4, dynamically controlling the energy storage system through the operating status evaluation attributes and the performance change trend.

[0010] On the basis of the above technical solution, preferably, step S1 includes the following steps:

[0011] S11, collecting original state evaluation data of the energy storage system, wherein the original state evaluation data includes original state data of each battery unit and original environmental data of the distribution station area;

[0012] S12, performing outlier detection, denoising, data standardization, and missing data processing on the original state assessment data to obtain state data and environmental data;

[0013] S13, fusing the state data and the environment data to obtain state assessment data.

[0014] Based on the above technical solution, preferably, step S11 includes:

[0015] The voltage data, current data, temperature data, SOC data, charge and discharge times of each battery cell, and the ambient temperature data, humidity data, load data, and grid voltage data of the distribution station area are collected at a preset collection frequency, and a preliminary check is performed on the collected data to mark abnormal data. The original state data includes voltage data, current data, temperature data, SOC data, and charge and discharge times. The original environmental data includes ambient temperature data, humidity data, load data, and grid voltage data. The original state data and original environmental data are stored in chronological order.

[0016] On the basis of the above technical solution, preferably, step S12 includes:

[0017] Identify the abnormality of the original state evaluation data by a statistical method, retain the original state evaluation data whose abnormality does not exceed the abnormality threshold, and obtain the first state evaluation data;

[0018] Performing denoising processing on the first state evaluation data by using a Gaussian filtering algorithm to obtain second state evaluation data;

[0019] Normalizing the second state evaluation data by Min-Max normalization to obtain third state evaluation data;

[0020] Missing values ​​are identified for the third state evaluation data to obtain a missing data type of the third state evaluation data, and mean interpolation, linear interpolation or multiple interpolation are used to fill in the third state evaluation data according to the missing data type to obtain state data and environment data, wherein the state data and environment data are respectively preprocessed original state data and original environment data.

[0021] On the basis of the above technical solution, preferably, step S13 includes:

[0022] The state data and the environment data are aligned on the time axis, and the first key feature and the second key feature are extracted from the state data and the environment data respectively. According to the time series characteristics of the first key feature and the second key feature, the first key feature and the second key feature are fused through a recursive fusion algorithm to obtain the state assessment data.

[0023] Based on the above technical solution, preferably, the calculation formula of the comprehensive evaluation model is:

[0024] CHI(x)=[ω1SOC,ω2T,ω3N,ω4SVM output (x)] T ;

[0025] SVM output (x) = sign(α i *y i *K(x i ,x)+b);

[0026] Among them, x is the state evaluation data, CHI(x) is the operating state evaluation attribute corresponding to the state evaluation data, ω1, ω2, ω3, and ω4 are the weight coefficients of SOC, temperature, battery cell charge and discharge times, and SVM output, respectively. SOC, T, N, and SVM output (x) are SOC, temperature, battery cell charge and discharge times, and SVM output, respectively. sign(·) is the signal function. α i is the Lagrange multiplier, y i is the category label, K(x i ,x) is the kernel function and b is the bias.

[0027] On the basis of the above technical solution, preferably, the performance prediction model construction process includes:

[0028] Arrange the state evaluation data in chronological order and divide it into training set and test set in a ratio of 7:3;

[0029] Constructing a long short-term memory network model, wherein the long short-term memory network model includes an input layer, a long short-term memory network layer, and an output layer;

[0030] The gradient value of the long short-term memory network model is calculated through the back-propagation algorithm, and the parameters of the long short-term memory network model are updated through the optimizer;

[0031] The LSTM model is trained using the training set until the loss function converges or the preset number of iterations is reached;

[0032] The long short-term memory network model is evaluated through the test set to obtain the performance prediction model.

[0033] Based on the above technical solution, preferably, step S4 includes:

[0034] S41, determining an operation mode and control strategy of the energy storage system according to the operation status evaluation attributes and performance change trends;

[0035] S42, performing real-time control on the energy storage system according to the determined operation mode and control strategy.

[0036] Based on the above technical solution, preferably, step S41 includes:

[0037] The current operating state of the energy storage system is judged according to the operating state evaluation attributes. The future performance changes of the energy storage system are predicted through the operating state evaluation attributes and the performance change trend. The future performance changes correspond to future time periods of a fixed period. A specified operating mode is selected based on the current operating state and the future performance changes. The specified operating mode includes a charging mode, a discharging mode and a standby mode. A specified control strategy is formulated according to the specified operating mode. The specified control strategy includes setting the charging and discharging power and controlling the charging and discharging time length.

[0038] Based on the above technical solution, preferably, step S42 includes:

[0039] Send an operation mode switching instruction to the energy storage system according to the specified operation mode, adjust the operation parameters of the energy storage system according to the specified control strategy, the operation parameters include charge and discharge current and charge and discharge voltage, monitor the current operation status of the energy storage system in real time, and record the control process data.

[0040] The method for controlling the operation of a flexible direct current energy storage system in a distribution station area of ​​the present invention has the following beneficial effects compared with the prior art:

[0041] (1) Through data collection, preprocessing, comprehensive evaluation and performance prediction, the original state data of the energy storage system is comprehensively analyzed and processed to obtain the operating state evaluation attributes and performance change trends, thereby realizing dynamic control, improving the adaptability of the energy storage system, avoiding overcharging and over-discharging, extending the battery life, and realizing more accurate and flexible energy storage system operation control;

[0042] (2) By performing outlier detection, denoising, data standardization, and missing data processing on the original status assessment data, high-quality status data and environmental data are obtained, and status assessment data is obtained through data fusion, thereby improving the accuracy of data processing;

[0043] (3) By building a performance prediction model and using the long short-term memory network to predict the performance change trend of the energy storage system, predictive management and dynamic control of the energy storage system can be achieved, thereby enhancing the adaptability of the system;

[0044] (4) Based on the operating status evaluation attributes and performance change trends, real-time control strategies are formulated and implemented, including switching of charging and discharging modes and parameter adjustments, thus achieving refined management and optimized operation of the energy storage system. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 The present invention is a flow chart of a method for controlling the operation of a flexible direct current energy storage system in a distribution station area. DETAILED DESCRIPTION

[0047] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] See also Figure 1 This embodiment provides a method for controlling the operation of a flexible direct current energy storage system in a distribution station area, comprising the following steps:

[0049] S1, collecting original state evaluation data of the energy storage system in operation, and preprocessing the original state evaluation data to obtain state evaluation data;

[0050] S2, identifying the state assessment data through the comprehensive assessment model to obtain the operating state assessment attributes of the energy storage system;

[0051] S3, identifying the state assessment data based on the performance prediction model to obtain the performance change trend of the energy storage system;

[0052] S4, dynamically controlling the energy storage system through the operating status evaluation attributes and the performance change trend.

[0053] Specifically, this embodiment comprehensively analyzes and processes the original state data of the energy storage system through data collection, preprocessing, comprehensive evaluation and performance prediction, and obtains the operating state evaluation attributes and performance change trends, thereby realizing dynamic control, improving the adaptability of the energy storage system, avoiding overcharging and over-discharging, extending the battery life, and realizing more accurate and flexible operation control of the energy storage system.

[0054] Step S1 includes the following steps:

[0055] S11, collecting original state evaluation data of the energy storage system, wherein the original state evaluation data includes original state data of each battery unit and original environmental data of the distribution station area;

[0056] S12, performing outlier detection, denoising, data standardization, and missing data processing on the original state assessment data to obtain state data and environmental data;

[0057] S13, fusing the state data and the environment data to obtain state assessment data.

[0058] Specifically, step S1 of this embodiment realizes systematic processing of energy storage system data by integrating raw data from data collection, preprocessing to data fusion, improves the consistency of overall data processing, comprehensively processes and integrates battery unit status data and distribution station area environmental data, and realizes multi-angle and all-round evaluation of the operating status of the energy storage system.

[0059] Step S11 includes:

[0060] The voltage data, current data, temperature data, SOC data, charge and discharge times of each battery cell, and the ambient temperature data, humidity data, load data, and grid voltage data of the distribution station area are collected at a preset collection frequency, and a preliminary check is performed on the collected data to mark abnormal data. The original state data includes voltage data, current data, temperature data, SOC data, and charge and discharge times. The original environmental data includes ambient temperature data, humidity data, load data, and grid voltage data. The original state data and original environmental data are stored in chronological order.

[0061] Specifically, step S11 of this embodiment realizes comprehensive monitoring of the energy storage system and its operating environment by collecting various parameters of the battery unit (voltage, current, temperature, SOC, charge and discharge times) and environmental data of the distribution station area (temperature, humidity, load, grid voltage).

[0062] By performing a preliminary check on the collected data and marking abnormal data, potential problems can be discovered in a timely manner. The original status data and original environmental data are stored in chronological order to facilitate subsequent time series analysis, which helps to identify the changing trends of system performance over time.

[0063] Data collection is performed by presetting the collection frequency, which can be adjusted according to actual needs to optimize system resource utilization while ensuring data quality.

[0064] Step S12 includes:

[0065] Identify the abnormality of the original state evaluation data by a statistical method, retain the original state evaluation data whose abnormality does not exceed the abnormality threshold, and obtain the first state evaluation data;

[0066] Performing denoising processing on the first state evaluation data by using a Gaussian filtering algorithm to obtain second state evaluation data;

[0067] Normalizing the second state evaluation data by Min-Max normalization to obtain third state evaluation data;

[0068] Missing values ​​are identified for the third state evaluation data to obtain a missing data type of the third state evaluation data, and mean interpolation, linear interpolation or multiple interpolation are used to fill in the third state evaluation data according to the missing data type to obtain state data and environment data, wherein the state data and environment data are respectively preprocessed original state data and original environment data.

[0069] Specifically, step S12 of this embodiment improves the quality and reliability of the data by identifying and removing outliers and performing Gaussian filtering and denoising, and uses Min-Max standardization to normalize the data so that data of different types and magnitudes can be uniformly compared and analyzed, thereby improving the accuracy of data processing.

[0070] Through missing value identification and a variety of interpolation methods (mean interpolation, linear interpolation, multiple interpolation), the problem of missing data is solved to ensure the integrity and continuity of the data. The corresponding interpolation method is used according to the different types of missing data, which reflects the flexibility and adaptability of the processing method and can better handle various types of missing data.

[0071] Step S13 includes:

[0072] The state data and the environment data are aligned on the time axis, and the first key feature and the second key feature are extracted from the state data and the environment data respectively. According to the time series characteristics of the first key feature and the second key feature, the first key feature and the second key feature are fused through a recursive fusion algorithm to obtain the state assessment data.

[0073] Specifically, step S13 of this embodiment ensures the time consistency of data from different sources by aligning the status data with the environmental data on the time axis, extracts key features from the status data and the environmental data respectively, realizes data dimensionality reduction and information concentration, and highlights the most important information for evaluating the operating status of the energy storage system.

[0074] The time series characteristics of the first key feature and the second key feature are used for fusion, which retains the temporal dynamic information of the data and helps to capture the performance trend of the energy storage system over time. The key features are fused through a recursive fusion algorithm to achieve an organic combination of state data and environmental data, and obtain more comprehensive and accurate state assessment data.

[0075] The calculation formula of the comprehensive evaluation model is:

[0076] CHI(x)=[ω1SOC,ω2T,ω3N,ω4SVM output (x)] T ;

[0077] SVM output (x) = sign(α i *y i *K(x i ,x)+b);

[0078] Among them, x is the state evaluation data, CHI(x) is the operating state evaluation attribute corresponding to the state evaluation data, ω1, ω2, ω3, and ω4 are the weight coefficients of SOC, temperature, battery cell charge and discharge times, and SVM output, respectively. SOC, T, N, and SVM output (x) are SOC, temperature, battery cell charge and discharge times, and SVM output, respectively. sign(·) is the signal function. α i is the Lagrange multiplier, y i is the category label, K(x i ,x) is the kernel function and b is the bias.

[0079] Specifically, the comprehensive evaluation model of this embodiment comprehensively considers multiple key factors such as SOC, temperature, charge and discharge times, and SVM output, and realizes a comprehensive evaluation of the operating status of the energy storage system. By introducing weight coefficients ω1, ω2, ω3, and ω4, the model can flexibly adjust the importance of each factor according to different application scenarios or operating conditions, thereby enhancing the adaptability of the evaluation model.

[0080] By introducing support vector machine (SVM) and kernel function, the model can effectively capture the nonlinear characteristics of the operating state of the energy storage system and improve the ability to identify complex operating states.

[0081] By integrating multiple factors into an operating status evaluation attribute CHI(x), the operating status of the energy storage system can be quantitatively represented, which facilitates subsequent decision analysis and control strategy formulation.

[0082] The SVM output is converted through the sign signal function to achieve the classification of the operating status. The classification results are combined with other continuous variables (such as SOC, temperature, etc.) to achieve an organic combination of discrete classification and continuous evaluation.

[0083] By introducing the Lagrange multiplier α i and bias b. The model incorporates the idea of ​​optimization theory and can find the optimal solution while satisfying the constraints, thus improving the performance and stability of the model.

[0084] The performance prediction model building process includes:

[0085] Arrange the state evaluation data in chronological order and divide it into training set and test set in a ratio of 7:3;

[0086] Constructing a long short-term memory network model, wherein the long short-term memory network model includes an input layer, a long short-term memory network layer, and an output layer;

[0087] The gradient value of the long short-term memory network model is calculated through the back-propagation algorithm, and the parameters of the long short-term memory network model are updated through the optimizer;

[0088] The LSTM model is trained using the training set until the loss function converges or the preset number of iterations is reached;

[0089] The long short-term memory network model is evaluated through the test set to obtain the performance prediction model.

[0090] Specifically, the performance prediction model of this embodiment effectively captures the dynamic characteristics of the energy storage system performance changing over time by arranging the state evaluation data in chronological order and using the long short-term memory network (LSTM) model, thereby improving the accuracy of the prediction.

[0091] The training set and test set are divided into a ratio of 7:3, and the model performance is evaluated by the test set, which effectively prevents the overfitting problem and enhances the generalization ability of the model, enabling it to adapt to different operating scenarios.

[0092] By dynamically updating the model parameters through the back-propagation algorithm and optimizer, the model's adaptive learning is achieved, and it can automatically adjust according to different data characteristics, thereby improving the model's adaptability and prediction accuracy.

[0093] The termination condition of loss function convergence or reaching the preset number of iterations is set, which not only ensures the adequacy of model training, but also avoids the waste of computing resources caused by over-training.

[0094] Step S4 includes:

[0095] S41, determining an operation mode and control strategy of the energy storage system according to the operation status evaluation attributes and performance change trends;

[0096] S42, performing real-time control on the energy storage system according to the determined operation mode and control strategy.

[0097] Specifically, step S4 of this embodiment realizes dynamic adjustment of the operation mode and control strategy of the energy storage system by comprehensively evaluating the operation status attributes and performance change trends. The adaptive control mechanism can flexibly respond according to the current status and predicted trends, thereby improving the operation efficiency of the system.

[0098] By adjusting specific operating parameters such as charging and discharging current and voltage in real time, refined control of the energy storage system is achieved, which not only improves system performance, but also helps to extend battery life and improve overall benefits.

[0099] By real-time monitoring of the system operating status and recording control process data, a complete closed-loop optimization process is formed, which not only ensures the safe and stable operation of the system, but also provides data support for performance analysis and strategy optimization, and realizes continuous improvement of the system.

[0100] Step S41 includes:

[0101] The current operating state of the energy storage system is judged according to the operating state evaluation attributes. The future performance changes of the energy storage system are predicted through the operating state evaluation attributes and the performance change trend. The future performance changes correspond to future time periods of a fixed period. A specified operating mode is selected based on the current operating state and the future performance changes. The specified operating mode includes a charging mode, a discharging mode and a standby mode. A specified control strategy is formulated according to the specified operating mode. The specified control strategy includes setting the charging and discharging power and controlling the charging and discharging time length.

[0102] Specifically, step S41 of this embodiment achieves multi-dimensional optimization decision-making on the operation mode of the energy storage system by comprehensively considering the current operation status and future performance changes, thereby improving the overall efficiency of the system operation.

[0103] By predicting future performance changes within a fixed period, predictive management of the energy storage system is achieved, which can respond to possible performance changes in advance and improve the stability of the system.

[0104] The charging, discharging or standby mode can be flexibly selected according to the current status and future trends, realizing the dynamic adjustment of the operation mode of the energy storage system and improving the system's adaptability to different operation scenarios.

[0105] By formulating specific control strategies including charging and discharging power and time length, refined management of the energy storage system is achieved, effectively improving energy utilization efficiency and system life.

[0106] Step S42 includes:

[0107] Send an operation mode switching instruction to the energy storage system according to the specified operation mode, adjust the operation parameters of the energy storage system according to the specified control strategy, the operation parameters include charge and discharge current and charge and discharge voltage, monitor the current operation status of the energy storage system in real time, and record the control process data.

[0108] Specifically, step S42 of this embodiment implements dynamic real-time control of the energy storage system by sending operation mode switching instructions and adjusting operation parameters in real time, thereby improving the system's response speed to environmental changes and demand changes.

[0109] By adjusting specific operating parameters such as charge and discharge current and voltage, refined control of the energy storage system is achieved, which helps to optimize system performance, extend battery life and improve energy utilization efficiency.

[0110] By real-time monitoring of the current operating status of the energy storage system, a closed-loop control-monitoring-optimization process is formed, which can promptly detect and handle abnormal conditions during operation and ensure the safe and stable operation of the system.

[0111] By recording control process data, data support is provided for system performance analysis, fault diagnosis and optimization strategy formulation, which contributes to the continuous improvement and long-term optimization of the system.

[0112] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for controlling the operation of a flexible direct current energy storage system in a distribution station area, characterized in that: The following steps are involved: S1, collecting original state evaluation data of the energy storage system in operation, and preprocessing the original state evaluation data to obtain state evaluation data; S2, identifying the state assessment data through the comprehensive assessment model to obtain the operating state assessment attributes of the energy storage system; S3, identifying the state assessment data based on the performance prediction model to obtain the performance change trend of the energy storage system; S4, dynamically controlling the energy storage system through the operating status evaluation attributes and the performance change trend.

2. The method for controlling the operation of a flexible direct current energy storage system in a distribution area according to claim 1, characterized in that: Step S1 includes the following steps: S11, collecting original state evaluation data of the energy storage system, wherein the original state evaluation data includes original state data of each battery unit and original environmental data of the distribution station area; S12, performing outlier detection, denoising, data standardization, and missing data processing on the original state assessment data to obtain state data and environmental data; S13, fusing the state data and the environment data to obtain state assessment data.

3. The method for controlling the operation of a flexible direct current energy storage system in a distribution area according to claim 2, characterized in that: Step S11 includes: The voltage data, current data, temperature data, SOC data, charge and discharge times of each battery cell, and the ambient temperature data, humidity data, load data, and grid voltage data of the distribution station area are collected at a preset collection frequency, and a preliminary check is performed on the collected data to mark abnormal data. The original state data includes voltage data, current data, temperature data, SOC data, and charge and discharge times. The original environmental data includes ambient temperature data, humidity data, load data, and grid voltage data. The original state data and original environmental data are stored in chronological order.

4. The method for controlling the operation of a flexible direct current energy storage system in a distribution area according to claim 3, characterized in that: Step S12 includes: Identify the abnormality of the original state evaluation data by a statistical method, retain the original state evaluation data whose abnormality does not exceed the abnormality threshold, and obtain the first state evaluation data; Performing denoising processing on the first state evaluation data by using a Gaussian filtering algorithm to obtain second state evaluation data; Normalizing the second state evaluation data by Min-Max normalization to obtain third state evaluation data; Missing values ​​are identified for the third state evaluation data to obtain a missing data type of the third state evaluation data, and mean interpolation, linear interpolation or multiple interpolation are used to fill in the third state evaluation data according to the missing data type to obtain state data and environment data, wherein the state data and environment data are respectively preprocessed original state data and original environment data.

5. The method for controlling the operation of a flexible direct current energy storage system in a distribution area according to claim 4, characterized in that: Step S13 includes: The state data and the environment data are aligned on the time axis, and the first key feature and the second key feature are extracted from the state data and the environment data respectively. According to the time series characteristics of the first key feature and the second key feature, the first key feature and the second key feature are fused through a recursive fusion algorithm to obtain the state assessment data.

6. The method for controlling the operation of a flexible direct current energy storage system in a distribution area according to claim 5, characterized in that: The calculation formula of the comprehensive evaluation model is: CHI(x)=[ω1SOC,ω2T,ω3N,ω4SVM output (x)] T ; SVM output (x)=sign(α i *y i *K(x i ,x)+b); Among them, x is the state evaluation data, CHI(x) is the operating state evaluation attribute corresponding to the state evaluation data, ω1, ω2, ω3, and ω4 are the weight coefficients of SOC, temperature, battery cell charge and discharge times, and SVM output, respectively. SOC, T, N, and SVM output (x) are SOC, temperature, battery cell charge and discharge times, and SVM output, respectively. sign(·) is the signal function. α i is the Lagrange multiplier, y i is the category label, K(x i ,x) is the kernel function and b is the bias.

7. The method for controlling the operation of a flexible direct current energy storage system in a distribution area according to claim 6, characterized in that: The performance prediction model building process includes: Arrange the state evaluation data in chronological order and divide it into training set and test set in a ratio of 7:3; Constructing a long short-term memory network model, wherein the long short-term memory network model includes an input layer, a long short-term memory network layer, and an output layer; The gradient value of the long short-term memory network model is calculated through the back-propagation algorithm, and the parameters of the long short-term memory network model are updated through the optimizer; The LSTM model is trained using the training set until the loss function converges or the preset number of iterations is reached; The long short-term memory network model is evaluated through the test set to obtain the performance prediction model.

8. The method for controlling the operation of a flexible direct current energy storage system in a distribution area according to claim 7, characterized in that: Step S4 includes: S41, determining an operation mode and control strategy of the energy storage system according to the operation status evaluation attributes and performance change trends; S42, performing real-time control on the energy storage system according to the determined operation mode and control strategy.

9. The method for controlling the operation of a flexible direct current energy storage system in a distribution area according to claim 8, characterized in that: Step S41 includes: The current operating state of the energy storage system is judged according to the operating state evaluation attributes. The future performance changes of the energy storage system are predicted through the operating state evaluation attributes and the performance change trend. The future performance changes correspond to future time periods of a fixed period. A specified operating mode is selected based on the current operating state and the future performance changes. The specified operating mode includes a charging mode, a discharging mode and a standby mode. A specified control strategy is formulated according to the specified operating mode. The specified control strategy includes setting the charging and discharging power and controlling the charging and discharging time length.

10. The method for controlling the operation of a flexible direct current energy storage system in a distribution area according to claim 9, characterized in that: Step S42 includes: Send an operation mode switching instruction to the energy storage system according to the specified operation mode, adjust the operation parameters of the energy storage system according to the specified control strategy, the operation parameters include charge and discharge current and charge and discharge voltage, monitor the current operation status of the energy storage system in real time, and record the control process data.

Images