Local power distribution method and system for energy storage power station

Abstract

The invention discloses a local power distribution method for an energy storage power station. The method comprises the following steps: acquiring and preprocessing data information of a battery module of a target energy storage power station; performing clustering analysis on the battery modules of the target energy storage power station based on a DBSCAN algorithm; according to a clustering analysis result, constructing a decision matrix according to the charge state, the health state and the residual capacity of the battery module, and calculating the priority of each clustered battery module cluster in combination with an entropy weight method; and constructing a local power distribution model of the target energy storage power station by taking the loss, the aging cost, the charge state balance degree and power distribution optimization based on the priority of the battery module as targets, solving the local power distribution model, and completing local power distribution of the target energy storage power station according to a solving result. The invention also discloses a system for realizing the local power distribution method of the energy storage power station. According to the method, local power distribution of the energy storage power station can be achieved, the characteristics of the battery modules are considered, and the scheme is higher in reliability and better in safety.

Description

Technical Field

[0001] The present invention belongs to the field of electrical automation, and in particular relates to a local power distribution method and system for an energy storage power station. Background Art

[0002] With the development of economy and technology and the improvement of people's living standards, electricity has become an indispensable secondary energy source in people's production and life, bringing endless convenience to people's production and life. Therefore, ensuring a stable and reliable supply of electricity has become one of the most important tasks of the power system.

[0003] Currently, more and more renewable energy power generation systems are being integrated into the power grid and generating electricity on a large scale. As an important means of regulating the intermittent and volatile nature of renewable energy, the safe and reliable operation of energy storage power stations is crucial to the power system.

[0004] Energy storage power stations are composed of numerous battery modules. Over long-term operation, these modules gradually develop significant differences in state of charge (SOC), state of health (SOH), and internal resistance due to subtle differences in manufacturing processes and environmental factors such as temperature, humidity, and charge / discharge frequency. Currently, power allocation schemes for energy storage power stations employ a traditional "one-size-fits-all" approach. This approach disregards individual differences in SOC, SOH, internal resistance, and other characteristics of each battery module, applying a uniform power allocation strategy to all modules. This approach fails to differentiate power levels for battery modules in different states, potentially leading to overuse of healthy modules due to their performance advantages, accelerating their degradation. Furthermore, it can pose safety risks by failing to consider the coupled relationship between SOC and internal resistance. Moreover, existing solutions often ignore the close coupling relationship between SOC and internal resistance. When the internal resistance of the battery module increases, if it is still charged and discharged according to conventional power, it is very easy to cause dangerous situations of overcharging and over-discharging of the battery. This not only seriously affects the service life of the battery, but may also bring safety hazards. Summary of the Invention

[0005] One of the objectives of the present invention is to provide a local power distribution method for an energy storage power station with high reliability and good safety.

[0006] A second object of the present invention is to provide a system for implementing the local power distribution method of the energy storage power station.

[0007] The local power distribution method of the energy storage power station provided by the present invention includes the following steps: S1. Obtain data information of the battery module of the target energy storage power station; S2 preprocesses the data information obtained in step S1; S3. Based on the preprocessed data information obtained in step S2, cluster analysis of the battery modules of the target energy storage power station is performed based on the DBSCAN algorithm; S4. Based on the cluster analysis results obtained in step S3, a decision matrix is ​​constructed based on the battery module's state of charge, health status, and remaining capacity. The priority of each clustered battery module cluster is calculated using the entropy weight method. S5. Build a local power allocation model for the target energy storage plant, optimizing power allocation based on losses, aging costs, state-of-charge balance, and battery module priority. S6. Solve the local power distribution model constructed in step S5, and complete the local power distribution of the target energy storage power station based on the solution result.

[0008] The step S1 specifically includes the following steps: Obtain data information of the battery module of the target energy storage power station; the data information includes state of charge SOC, health state SOH and internal resistance .

[0009] The step S2 specifically includes the following steps: The preprocessing process includes data cleaning, data interpolation and data normalization; The data normalization is specifically performed using the following formula: , where is the data after normalization; X is the data before normalization; is the minimum value of the data; is the maximum value of the data.

[0010] The step S3 comprises the following steps: The pre-processed state of charge, health state and internal resistance are used as three-dimensional feature vectors to construct the feature space; The DBSCAN algorithm is used to perform cluster analysis on the battery modules of the target energy storage power station: During cluster analysis, weighted Euclidean distance is used to represent the similarity of each sample in the feature space; the neighborhood radius during clustering is adjusted based on the temperature attenuation effect on performance and the stability of the battery module health status; and the minimum number of samples for each cluster is set based on the total number of battery modules.

[0011] The step S3 specifically includes the following steps: The similarity between any two samples x and y in the feature space is calculated using the following formula: , where is the similarity between sample x and sample y; is the SOC value of sample x; is the SOC value of sample y; is the SOH value of sample x; is the SOH value of sample y; is the internal resistance value of sample x; is the internal resistance value of sample y; is the set SOC value weight, is the set SOH value weight, is the set internal resistance weight, and ; Based on the temperature attenuation effect on performance and the stability of the battery module health status, the following formula is used to adjust the neighborhood radius of clustering in real time: : , where is the set base radius value; is the set health fluctuation weight value; is a stability indicator of the battery module's health status, and , is the standard deviation calculation function, The health status (SOH) historical data of the battery module in the past three days; Correction weight value for the set temperature; is the temperature correction coefficient of the battery module, and , is the setting parameter for temperature correction, is the maximum temperature of the battery module; According to the total number of battery modules, the minimum number of samples for each cluster is calculated using the following formula: , where is the minimum number of samples for each cluster; is the total number of battery modules; This is a round-up operation; When clustering, for each sample point p, if the neighborhood radius of p The number of samples in the sample is not less than the minimum number of samples , then the sample point p is used as the cluster core; if the neighborhood radius of the cluster core p If there are other cluster cores q, then cluster core q is added to the cluster cluster corresponding to cluster core p, and clustering is continued with cluster core q as the center; Finally, clustering is completed.

[0012] The step S4 comprises the following steps: A decision matrix is ​​constructed using the battery module's state of charge (SOC), state of health (SOH), and remaining capacity as indicators. Standardize the constructed decision matrix; According to the standardized decision matrix, the entropy weight method is used to calculate the entropy value and corresponding weight value of each indicator; According to the weight value of each indicator and the constructed decision matrix, the priority of each clustered battery module cluster is calculated.

[0013] The step S4 specifically includes the following steps: The decision matrix is ​​constructed using the state of charge (SOC), state of health (SOH), and remaining capacity of the clustered battery module cluster as indicators. Among them, the state of charge (SOC) of the battery module cluster is the arithmetic mean of the state of charge (SOC) of all battery modules in the cluster; the state of health (SOH) of the battery module cluster is the arithmetic mean of the state of health (SOH) of all battery modules in the cluster; the remaining capacity of the battery module cluster is the sum of the remaining capacities of all battery modules in the cluster; and the internal resistance of the battery module cluster is the sum of the internal resistances of all battery modules in the cluster. The constructed decision matrix is ​​standardized to obtain a standardized decision matrix Q; Based on the entropy weight method, the entropy value and corresponding weight value of each indicator are calculated using the following formula: , , where is the entropy value of the jth indicator; m is the total number of battery module clusters; is the element in row i and column j of the standardized decision matrix Q; is the weight value of the jth indicator; Based on the weight values ​​of each indicator and the constructed decision matrix, the priority of each clustered battery module cluster is calculated using the following formula: , where is the priority of the i-th battery module cluster.

[0014] The step S5 comprises the following steps: Calculating a priority weight factor for each battery module cluster based on the priority of each battery module cluster, and optimizing power allocation based on the priority of each battery module cluster calculated based on the priority weight factor of each battery module cluster; Taking the power allocation optimization of battery module cluster loss, aging cost, state of charge balance and priority as the objective function, and total power balance, power limit, SOC safety range and SOH safety range as constraints, a local power allocation model of the target energy storage power station is constructed.

[0015] The step S5 specifically includes the following steps: The following formula is used as the objective function of the local power distribution model of the target energy storage power station: , where is the power value of the i-th battery module cluster; is the equivalent internal resistance of the i-th battery module cluster; is the SOH value of the i-th battery module cluster; is the standard deviation of the SOC values ​​of the battery modules contained in the i-th battery module cluster; is the priority weight factor of the i-th battery module cluster, and , Indicates the maximum priority of the battery module cluster; is the set aging cost weight value; is the set state of charge balance weight value; Optimizing weight values ​​for power allocation of set priorities; The following formula is used as the constraint condition of the local power distribution model of the target energy storage power station: The following formula is used as the total power balance constraint: , where is the power regulation command value of the target energy storage power station; The following formula is used as the power limit constraint: , where is the maximum allowable power of the set i-th battery module cluster; The following formula is used as the SOC safety range constraint: , where is the minimum SOC value of the set i-th battery module cluster; is the maximum SOC value of the set i-th battery module cluster; is the duration of the control cycle; is the total capacity of the battery modules of the i-th battery module cluster; The following formula is used as the SOH safety range constraint: , where The critical health threshold of the battery module cluster is set.

[0016] Use the following formula to set the aging cost weight value : , where The base value of the set aging cost weight value; represents the decay rate of the SOH value of the i-th battery module cluster; The maximum value of the SOH decay rate of the battery module cluster.

[0017] The present invention also provides a system for implementing the local power distribution method of the energy storage power station, comprising a data acquisition module, a data processing module, a battery clustering module, a priority calculation module, a distribution modeling module and a power distribution module; the data acquisition module, the data processing module, the battery clustering module, the priority calculation module, the distribution modeling module and the power distribution module are connected in series in sequence; the data acquisition module is used to acquire data information of the battery modules of the target energy storage power station and upload the data information to the data processing module; the data processing module is used to pre-process the acquired data information according to the received data information and upload the data information to the battery clustering module; the battery clustering module is used to perform clustering of the battery modules of the target energy storage power station based on the received data information and the obtained pre-processed data information based on the DBSCAN algorithm Cluster analysis, and upload the data information to the priority calculation module; the priority calculation module is used to construct a decision matrix based on the received data information and the obtained cluster analysis results, with the state of charge, health status and remaining capacity of the battery module, and calculate the priority of each clustered battery module cluster in combination with the entropy weight method, and upload the data information to the allocation modeling module; the allocation modeling module is used to construct a local power distribution model of the target energy storage power station based on the received data information, with the loss, aging cost, state of charge balance and power distribution optimization based on the priority of the battery module as the goal, and upload the data information to the power distribution module; the power distribution module is used to solve the constructed local power distribution model according to the received data information, and complete the local power distribution of the target energy storage power station according to the solution result.

[0018] The local power distribution method and system for an energy storage power station provided by the present invention clusters the battery modules by acquiring data information of the battery modules of the energy storage power station, calculates the priority of the battery modules based on the clustering results, and constructs and solves a local power distribution model of the target energy storage power station with the goals of optimizing power distribution based on loss, aging cost, state of charge balance, and priority of the battery modules. This not only realizes the local power distribution of the energy storage power station, but also takes into account the inherent characteristics of the battery modules, resulting in a more reliable and safer solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the process of the present invention.

[0020] Figure 2 Schematic diagram for comparing charge and discharge efficiency curves of embodiments of the method of the present invention.

[0021] Figure 3 Schematic diagram of the comparison of SOC standard deviation curves of the method embodiment of the present invention.

[0022] Figure 4Schematic diagram of SOH attenuation curve comparison of the embodiment of the method of the present invention.

[0023] Figure 5 Schematic diagram of the functional modules of the system of the present invention. DETAILED DESCRIPTION

[0024] like Figure 1 The figure shows a flow chart of the method of the present invention: The local power distribution method of the energy storage power station disclosed in the present invention includes the following steps: S1. Obtaining data information about the battery module of the target energy storage power station; specifically comprising the following steps: Obtain data information of the battery module of the target energy storage power station; the data information includes state of charge SOC, health state SOH and internal resistance .

[0025] S2 preprocesses the data information obtained in step S1; specifically comprises the following steps: The preprocessing process includes data cleaning, data interpolation and data normalization; Since the dimensions of the acquired data are different, in order to eliminate the impact of dimensional differences on subsequent analysis, the following formula is used for normalization: , where is the data after normalization; X is the data before normalization; is the minimum value of the data; is the maximum value of the data; Through this processing step, data of different dimensions are uniformly mapped to interval, laying a good foundation for subsequent data analysis and processing.

[0026] S3. Based on the pre-processed data information obtained in step S2, cluster analysis of the battery modules of the target energy storage power station is performed based on the DBSCAN algorithm; comprising the following steps: The pre-processed state of charge, health status and internal resistance are used as three-dimensional feature vectors to construct a feature space to fully reflect the status of each battery module; The DBSCAN algorithm is used to perform cluster analysis on the battery modules of the target energy storage power station: During cluster analysis, weighted Euclidean distance is used to represent the similarity of each sample in the feature space; the neighborhood radius during clustering is adjusted based on the temperature attenuation effect on performance and the stability of the battery module health status; and the minimum number of samples for each cluster is set based on the total number of battery modules.

[0027] The following steps can be taken for implementation: The similarity between any two samples x and y in the feature space is calculated using the following formula: , where is the similarity between sample x and sample y; is the SOC value of sample x; is the SOC value of sample y; is the SOH value of sample x; is the SOH value of sample y; is the internal resistance value of sample x; is the internal resistance value of sample y; is the set SOC value weight, is the set SOH value weight, is the set internal resistance weight, and ; When implementing, 、 and The value of can be set by researchers. Based on the temperature attenuation effect on performance and the stability of the battery module health status, the following formula is used to adjust the neighborhood radius of clustering in real time: : , where is the set base radius value; is the set health fluctuation weight value; is a stability indicator of the battery module's health status, and , is the standard deviation calculation function, The health status (SOH) historical data of the battery module in the past three days; Correction weight value for the set temperature; is the temperature correction coefficient of the battery module, and , is the setting parameter for temperature correction, is the maximum temperature of the battery module; According to the total number of battery modules, the minimum number of samples for each cluster is calculated using the following formula: , where is the minimum number of samples for each cluster; is the total number of battery modules; This is a round-up operation; When clustering, for each sample point p, if the neighborhood radius of p The number of samples in the sample is not less than the minimum number of samples , then the sample point p is used as the cluster core; if the neighborhood radius of the cluster core p If there are other cluster cores q, then cluster core q is added to the cluster cluster corresponding to cluster core p, and clustering is continued with cluster core q as the center; Finally, clustering is completed.

[0028] After clustering is completed, there may be some points in the feature space that are not assigned to any cluster. The battery modules corresponding to these points can be identified as abnormal modules, such as those with poor health status, significantly different state of charge from other battery modules, or abnormal internal resistance. These battery modules can be marked as abnormal and then subsequently inspected or replaced. Through the above improved clustering process, noise points in the data can be automatically identified, and the battery modules can be accurately divided into several groups, such as high health and high SOC group, high health and low SOC group, low health and high SOC group, low health and low SOC group, etc.; for each cluster, targeted control strategies can be formulated.

[0029] The core significance of clustering is: Simplify control complexity: By grouping modules with similar states into a cluster, the control object is reduced from a "single module" to a "cluster," reducing the number of optimization variables and ensuring that the power allocation model can be solved efficiently. Achieve differentiated regulation: After clustering, precise strategies can be formulated based on the characteristics of different clusters (such as high-health, high-SOC clusters and low-health, low-SOC clusters), avoiding overuse of high-health modules while reducing the "barrel effect" through intra-cluster balancing. Abnormal detection and management: Clustering can automatically identify abnormal modules (such as modules with SOH < 60% or abnormal internal resistance), facilitating individual tagging and maintenance, thereby improving system safety.

[0030] S4. Based on the cluster analysis results obtained in step S3, a decision matrix is ​​constructed based on the battery module's state of charge, health status, and remaining capacity. The priority of each clustered battery module cluster is calculated using the entropy weight method. The steps include: A decision matrix is ​​constructed using the battery module's state of charge (SOC), state of health (SOH), and remaining capacity as indicators. Standardize the constructed decision matrix; According to the standardized decision matrix, the entropy weight method is used to calculate the entropy value and corresponding weight value of each indicator; According to the weight value of each indicator and the constructed decision matrix, the priority of each clustered battery module cluster is calculated.

[0031] The following steps can be taken for implementation: The decision matrix is ​​constructed using the state of charge (SOC), state of health (SOH), and remaining capacity of the clustered battery module cluster as indicators. The state of charge (SOC) of a battery module cluster is the arithmetic mean of the SOCs of all the battery modules in the cluster, reflecting the overall charge level of the cluster. The state of health (SOH) of a battery module cluster is the arithmetic mean of the SOHs of all the battery modules in the cluster, representing the overall health of the cluster. The remaining capacity of a battery module cluster is the sum of the remaining capacities of all the battery modules in the cluster, used to measure the total energy storage capacity of the cluster. The internal resistance of a battery module cluster is the sum of the internal resistances of all the battery modules in the cluster, reflecting the overall resistance characteristics of the cluster. The constructed decision matrix is ​​standardized to obtain a standardized decision matrix Q; Based on the entropy weight method, the entropy value and corresponding weight value of each indicator are calculated using the following formula: , , where is the entropy value of the jth indicator; m is the total number of battery module clusters; is the element in row i and column j of the standardized decision matrix Q; is the weight value of the jth indicator; Based on the weight values ​​of each indicator and the constructed decision matrix, the priority of each clustered battery module cluster is calculated using the following formula: , where is the priority of the i-th battery module cluster.

[0032] Through the above priority calculation, the control priority of each module cluster can be objectively quantified based on the battery status data, and a dynamic and differentiated power allocation strategy can be implemented. Its core value lies in: ① Integrating the objective weighting of the entropy weight method to adapt to the dynamic changes of the battery status and avoid subjective decision-making bias; ② Coordinating multiple objectives such as life, efficiency, safety, and balance, optimizing energy distribution through priority sorting, which not only improves system performance but also delays battery aging; ③ Combining clustering to simplify control complexity, ensure real-time and operability in large-scale energy storage scenarios, provide quantitative decision-making support for the refined management of energy storage systems, and ultimately significantly enhance the system's reliability, economy and full life cycle operation efficiency.

[0033] S5. Build a local power allocation model for the target energy storage plant, optimizing power allocation based on losses, aging costs, state-of-charge balance, and battery module priority. This includes the following steps: Calculating a priority weight factor for each battery module cluster based on the priority of each battery module cluster, and optimizing power allocation based on the priority of each battery module cluster calculated based on the priority weight factor of each battery module cluster; Taking the power allocation optimization of battery module cluster loss, aging cost, state of charge balance and priority as the objective function, and total power balance, power limit, SOC safety range and SOH safety range as constraints, a local power allocation model of the target energy storage power station is constructed.

[0034] The following steps can be taken for implementation: The following formula is used as the objective function of the local power distribution model of the target energy storage power station: , where is the power value of the i-th battery module cluster; is the equivalent internal resistance of the i-th battery module cluster; is the SOH value of the i-th battery module cluster; is the standard deviation of the SOC values ​​of the battery modules contained in the i-th battery module cluster; is the priority weight factor of the i-th battery module cluster, and , Indicates the maximum priority of the battery module cluster; is the set aging cost weight value (preferably 0.7, which can be set by yourself or using the subsequent solution); is the set state of charge balance weight value (preferably 0.3); Optimizing weight values ​​for power allocation of set priorities; In the objective function, It is the loss item of the battery module cluster, which can reflect the loss of electrical energy converted into heat energy; It is an aging cost item that can reflect the aging status of the battery module cluster; The state of charge balance term reflects the difference in SOC values ​​within the cluster. This term is added to prevent capacity waste caused by the "barrel effect"; This is a priority power allocation optimization item. The purpose of adding this item is to make the battery module cluster with high priority bear more power allocation results.

[0035] The following formula is used as the constraint condition of the local power distribution model of the target energy storage power station: The following formula is used as the total power balance constraint: , where is the power regulation command value of the target energy storage power station; The following formula is used as the power limit constraint: , where is the maximum allowable power of the set i-th battery module cluster; The following formula is used as the SOC safety range constraint: , where is the minimum SOC value of the set i-th battery module cluster (preferably 5%); is the maximum SOC value of the set i-th battery module cluster (preferably 95%); is the duration of the control cycle; is the total capacity of the battery modules of the i-th battery module cluster; The following formula is used as the SOH safety range constraint: , where is the critical health threshold of the battery module cluster (preferably 60%); this constraint is used to ensure that the overuse of aged battery module clusters is avoided.

[0036] In addition, for the set aging cost weight value , can be calculated using the following formula: , where The base value of the set aging cost weight value; represents the decay rate of the SOH value of the i-th battery module cluster; The maximum value of the SOH decay rate of the battery module cluster.

[0037] S6. Solve the local power allocation model constructed in step S5, and complete the local power allocation of the target energy storage power station based on the solution result; in specific implementation, a quadratic programming scheme can be used to solve the model.

[0038] The method of the present invention is further described below with reference to an embodiment: Experimental setup: Dataset: A 100kWh energy storage system, consisting of 10 battery module clusters, simulates different SOH and SOC states; initial SOC range 30%-70%, SOH range 60%-95%, internal resistance 10-30 ; The comparison objects are: the solution of the present invention and the traditional method; among them, the traditional method is a unified charging and discharging solution for all battery modules, which does not distinguish between states and evenly distributes power, and ignores the individual differences of battery modules.

[0039] Evaluation indicators: charge and discharge efficiency, SOC standard deviation, SOH decay rate and number of overcharge and overdischarge times.

[0040] Experimental results: After 100 charge and discharge cycles, the final evaluation indicators are shown in Table 1: Table 1 Evaluation index table

[0041] As can be seen from Table 1, the method of the present invention is significantly superior to the traditional method in all evaluation indicators: in terms of charge and discharge efficiency, the method of the present invention reaches 92.5%, which is much higher than the 85.3% of the traditional method, indicating that the energy conversion efficiency of the scheme of the present invention is higher; in terms of SOC standard deviation, the method of the present invention is only 3.2%, which is much lower than the 8.7% of the traditional method, indicating that the present invention can better achieve the balance of the state of charge of the battery module; in terms of SOH decay rate, the method of the present invention is 1.2% / 100 times, which is significantly lower than the 3.5% / 100 times of the traditional method, reflecting that the present invention can effectively slow down the aging speed of the battery module; in terms of safety, the method of the present invention did not have overcharge and over-discharge phenomena, while the traditional method had 12 overcharge and over-discharge phenomena, which fully demonstrates that the safety of the method of the present invention is better. Overall, by considering the individual characteristics of the battery module and performing differentiated regulation, the method of the present invention has shown obvious advantages in improving efficiency, balance, delaying aging and ensuring safety.

[0042] Comparison curves such as Figures 2 to 4 As shown: pass Figure 2 It can be seen that the efficiency of the solution of the present invention is stable at 92%-93%, while the efficiency of the comparative solution fluctuates greatly (85%-87%); this indicates that the frequent participation of high internal resistance clusters in the comparative solution leads to increased losses.

[0043] pass Figure 3 It can be seen that the SOC standard deviation of the solution of the present invention quickly dropped to 3% and maintained, while the standard deviation of the comparison solution was always >8%; this shows that the comparison solution did not perform differentiated regulation, the high SOC cluster was over-discharged, and the low SOC cluster was under-charged, forming a "barrel effect".

[0044] pass Figure 4 It can be seen that the SOH of the solution of the present invention decays slowly, while the SOH of the comparative solution decays quickly; this indicates that the low SOH clusters in the comparative solution are overused, and the battery module ages faster.

[0045] like Figure 5The figure shows a functional module diagram of the system of the present invention: the system disclosed by the present invention for realizing the local power distribution method of the energy storage power station comprises a data acquisition module, a data processing module, a battery clustering module, a priority calculation module, a distribution modeling module and a power distribution module; the data acquisition module, the data processing module, the battery clustering module, the priority calculation module, the distribution modeling module and the power distribution module are connected in series in sequence; the data acquisition module is used to acquire the data information of the battery module of the target energy storage power station and upload the data information to the data processing module; the data processing module is used to pre-process the acquired data information according to the received data information and upload the data information to the battery clustering module; the battery clustering module is used to cluster the target energy storage power station according to the received data information and the obtained pre-processed data information based on the DBSCAN algorithm. The battery modules of the station are clustered and analyzed, and the data information is uploaded to the priority calculation module; the priority calculation module is used to construct a decision matrix based on the received data information and the obtained cluster analysis results, with the state of charge, health status and remaining capacity of the battery module, and calculate the priority of each clustered battery module cluster in combination with the entropy weight method, and upload the data information to the allocation modeling module; the allocation modeling module is used to construct a local power distribution model of the target energy storage power station based on the received data information, with the loss, aging cost, state of charge balance and power distribution optimization based on the priority of the battery module as the goal, and upload the data information to the power distribution module; the power distribution module is used to solve the constructed local power distribution model based on the received data information, and complete the local power distribution of the target energy storage power station according to the solution result.

Claims

1. A local power distribution method for an energy storage power station, characterized in that The steps include: S1. Obtain data information of the battery module of the target energy storage power station; S2 preprocesses the data information obtained in step S1; S3. Based on the preprocessed data information obtained in step S2, cluster analysis of the battery modules of the target energy storage power station is performed based on the DBSCAN algorithm; S4. Based on the cluster analysis results obtained in step S3, a decision matrix is ​​constructed based on the battery module's state of charge, health status, and remaining capacity. The priority of each clustered battery module cluster is calculated using the entropy weight method. S5. Build a local power allocation model for the target energy storage plant, optimizing power allocation based on losses, aging costs, state-of-charge balance, and battery module priority. S6. Solve the local power distribution model constructed in step S5, and complete the local power distribution of the target energy storage power station based on the solution result.

2. The local power distribution method of the energy storage power station according to claim 1 is characterized in that The step S1 specifically includes the following steps: Obtain data information of the battery module of the target energy storage power station; the data information includes state of charge SOC, health state SOH and internal resistance ; The step S2 specifically includes the following steps: The preprocessing process includes data cleaning, data interpolation and data normalization; The data normalization is specifically performed using the following formula: , where is the data after normalization; X is the data before normalization; is the minimum value of the data; is the maximum value of the data.

3. The local power distribution method of the energy storage power station according to claim 2, characterized in that The step S3 comprises the following steps: The pre-processed state of charge, health state and internal resistance are used as three-dimensional feature vectors to construct the feature space; The DBSCAN algorithm is used to perform cluster analysis on the battery modules of the target energy storage power station: During cluster analysis, weighted Euclidean distance is used to represent the similarity of each sample in the feature space; the neighborhood radius during clustering is adjusted based on the temperature attenuation effect on performance and the stability of the battery module health status; and the minimum number of samples for each cluster is set based on the total number of battery modules.

4. The local power distribution method of the energy storage power station according to claim 3 is characterized in that The step S3 specifically includes the following steps: The similarity between any two samples x and y in the feature space is calculated using the following formula: , where is the similarity between sample x and sample y; is the SOC value of sample x; is the SOC value of sample y; is the SOH value of sample x; is the SOH value of sample y; is the internal resistance value of sample x; is the internal resistance value of sample y; is the set SOC value weight, is the set SOH value weight, is the set internal resistance weight, and ; Based on the temperature attenuation effect on performance and the stability of the battery module health status, the following formula is used to adjust the neighborhood radius of clustering in real time: : , where is the set base radius value; is the set health fluctuation weight value; is a stability indicator of the battery module's health status, and , is the standard deviation calculation function, The health status (SOH) historical data of the battery module in the past three days; Correction weight value for the set temperature; is the temperature correction coefficient of the battery module, and , is the setting parameter for temperature correction, is the maximum temperature of the battery module; According to the total number of battery modules, the minimum number of samples for each cluster is calculated using the following formula: , where is the minimum number of samples for each cluster; is the total number of battery modules; This is a round-up operation; When clustering, for each sample point p, if the neighborhood radius of p The number of samples in the sample is not less than the minimum number of samples , then the sample point p is used as the clustering core; If the neighborhood radius of the cluster core p If there are other cluster cores q, then cluster core q is added to the cluster cluster corresponding to cluster core p, and clustering is continued with cluster core q as the center; Finally, clustering is completed.

5. The local power distribution method of the energy storage power station according to claim 4 is characterized in that The step S4 comprises the following steps: A decision matrix is ​​constructed using the battery module's state of charge (SOC), state of health (SOH), and remaining capacity as indicators. Standardize the constructed decision matrix; According to the standardized decision matrix, the entropy weight method is used to calculate the entropy value and corresponding weight value of each indicator; According to the weight value of each indicator and the constructed decision matrix, the priority of each clustered battery module cluster is calculated.

6. The local power distribution method of the energy storage power station according to claim 5, characterized in that The step S4 specifically includes the following steps: The decision matrix is ​​constructed using the state of charge (SOC), state of health (SOH), and remaining capacity of the clustered battery module cluster as indicators. Among them, the state of charge (SOC) of the battery module cluster is the arithmetic mean of the state of charge (SOC) of all battery modules in the cluster; the state of health (SOH) of the battery module cluster is the arithmetic mean of the state of health (SOH) of all battery modules in the cluster; the remaining capacity of the battery module cluster is the sum of the remaining capacities of all battery modules in the cluster; and the internal resistance of the battery module cluster is the sum of the internal resistances of all battery modules in the cluster. The constructed decision matrix is ​​standardized to obtain a standardized decision matrix Q; Based on the entropy weight method, the entropy value and corresponding weight value of each indicator are calculated using the following formula: , , where is the entropy value of the jth indicator; m is the total number of battery module clusters; is the element in row i and column j of the standardized decision matrix Q; is the weight value of the jth indicator; Based on the weight values ​​of each indicator and the constructed decision matrix, the priority of each clustered battery module cluster is calculated using the following formula: , where is the priority of the i-th battery module cluster.

7. The local power distribution method of the energy storage power station according to claim 6, characterized in that The step S5 comprises the following steps: Calculating a priority weight factor for each battery module cluster based on the priority of each battery module cluster, and optimizing power allocation based on the priority of each battery module cluster calculated based on the priority weight factor of each battery module cluster; Taking the power allocation optimization of battery module cluster loss, aging cost, state of charge balance and priority as the objective function, and total power balance, power limit, SOC safety range and SOH safety range as constraints, a local power allocation model of the target energy storage power station is constructed.

8. The local power distribution method of the energy storage power station according to claim 7, characterized in that The step S5 specifically includes the following steps: The following formula is used as the objective function of the local power distribution model of the target energy storage power station: , where is the power value of the i-th battery module cluster; is the equivalent internal resistance of the i-th battery module cluster; is the SOH value of the i-th battery module cluster; is the standard deviation of the SOC values ​​of the battery modules contained in the i-th battery module cluster; is the priority weight factor of the i-th battery module cluster, and , Indicates the maximum priority of the battery module cluster; is the set aging cost weight value; is the set state of charge balance weight value; Optimizing weight values ​​for power allocation of set priorities; The following formula is used as the constraint condition of the local power distribution model of the target energy storage power station: The following formula is used as the total power balance constraint: , where is the power regulation command value of the target energy storage power station; The following formula is used as the power limit constraint: , where is the maximum allowable power of the set i-th battery module cluster; The following formula is used as the SOC safety range constraint: , where is the minimum SOC value of the set i-th battery module cluster; is the maximum SOC value of the set i-th battery module cluster; is the duration of the control cycle; is the total capacity of the battery modules of the i-th battery module cluster; The following formula is used as the SOH safety range constraint: , where The critical health threshold of the battery module cluster is set.

9. The local power distribution method of the energy storage power station according to claim 8, characterized in that Use the following formula to set the aging cost weight value : , where The base value of the set aging cost weight value; represents the decay rate of the SOH value of the i-th battery module cluster; The maximum value of the SOH decay rate of the battery module cluster.

10. A system for implementing the local power distribution method of an energy storage power station according to any one of claims 1 to 9, characterized in that It includes a data acquisition module, a data processing module, a battery clustering module, a priority calculation module, a distribution modeling module, and a power distribution module; the data acquisition module, the data processing module, the battery clustering module, the priority calculation module, the distribution modeling module, and the power distribution module are connected in series in sequence; the data acquisition module is used to obtain data information of the battery modules of the target energy storage power station and upload the data information to the data processing module; The data processing module is used to pre-process the acquired data information according to the received data information, and upload the data information to the battery clustering module; the battery clustering module is used to perform cluster analysis on the battery modules of the target energy storage power station based on the received data information and the obtained pre-processed data information based on the DBSCAN algorithm, and upload the data information to the priority calculation module; The priority calculation module is used to construct a decision matrix based on the received data information and the cluster analysis results, using the battery module's state of charge, health status, and remaining capacity. It also uses the entropy weight method to calculate the priority of each clustered battery module cluster and upload the data information to the allocation modeling module. The distribution modeling module is used to construct a local power distribution model for the target energy storage power station based on the received data information, with the goals of power distribution optimization based on loss, aging cost, charge state balance, and battery module priority, and upload the data information to the power distribution module; the power distribution module is used to solve the constructed local power distribution model based on the received data information and complete the local power distribution of the target energy storage power station based on the solution results.

Images