Electrochemical energy storage power station online safety assessment method and device based on data driving

By constructing a battery safety assessment index system with multi-information fusion and game theory combination empowerment-TOPSIS model-driven method, combined with LSSVM network training, the rapid and accurate safety assessment of large-capacity electrochemical energy storage power station batteries is achieved, and the battery safety assessment is solved Problems with large amount of calculation and difficult data acquisition.

CN118981586APending Publication Date: 2024-11-19山东智源电力设计咨询有限公司 +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411101016.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-12
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The battery safety assessment of large-capacity electrochemical energy storage power stations during the operation stage has problems such as large calculation volume and limited rapidity, and the safety score data of the training samples are difficult to obtain through measurement.

Method used

Adopting the online safety evaluation method of electrochemical energy storage power stations based on data-driven, the battery cell is realized by constructing a multi-level safety evaluation index system with multi-level information fusion, combining game theory combination empowerment-TOPSIS model driving method and LSSVM network training, the battery cell is realized, Real-time safety assessment of modules and cabins.

Benefits of technology

It realizes rapid and accurate safety assessment of large-capacity electrochemical energy storage power station batteries, is suitable for health monitoring and early warning, solves the problem of obtaining training sample data, and has both accuracy and rapidity of evaluation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118981586A_ABST
    Figure CN118981586A_ABST
Patent Text Reader

Abstract

The invention discloses an electrochemical energy storage power station online safety assessment method and device based on data driving. The method comprises the following steps: constructing a multi-information fusion electrochemical energy storage power station battery multistage safety assessment index system; collecting historical monitoring data and calculating a safety evaluation index value; taking the safety evaluation index value as an input data sample, obtaining an output data sample consisting of safety scores of the battery monomers, the battery module and the battery cabin by adopting a model driving battery safety evaluation method based on game theory combination weighting-TOPSIS, and forming training sample data; performing network training by using the training sample data to obtain a battery safety assessment LSSVM model and parameters thereof; and obtaining a real-time safety assessment index value as the input of the model, and obtaining a battery monomer safety score, a battery module safety score and a battery cabin online safety assessment result. The problem that the safety score data of the training sample is difficult to obtain through a measurement mode is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a data-driven online safety assessment method and device for an electrochemical energy storage power station, belonging to the technical field of energy system safety situation awareness. Background Art

[0002] Electrochemical energy storage power stations have become a key technological support for new power systems, as they effectively balance electricity supply and demand and enhance the ability of new energy systems to absorb renewable energy. Lithium-ion batteries, with their low cost, high energy density, and environmental friendliness, are widely used in electrochemical energy storage power stations. However, lithium-ion batteries use flammable organic electrolytes with low boiling points. Failure of the batteries or electrical equipment can easily lead to thermal runaway, which can lead to major safety incidents such as fires and explosions in the entire energy storage system. Safety issues have become a primary obstacle to the large-scale application of electrochemical energy storage power stations, and in recent years, their safety issues have received increasing attention and research.

[0003] Safety control in electrochemical energy storage power stations encompasses multiple key aspects of the entire life cycle, including battery manufacturing, power station design and construction, operation and maintenance, and emergency response. Large-capacity electrochemical energy storage power stations mostly utilize prefabricated cabins, with the battery cabin forming the core component of the system. The cabin consists of a series-parallel arrangement of battery cells, battery modules, battery stacks, and battery clusters. The cabin houses a large number of battery cells. When a battery cell experiences thermal runaway, the heat released propagates to the battery modules, causing a cascading thermal runaway event that could potentially cause an accident in the entire energy storage system. Risk sources in electrochemical energy storage power stations include the battery itself, the operating environment, external shocks, and the management system. This results in a wide variety of precursors to thermal runaway. Consequently, battery safety assessments during the operational phase of large-capacity electrochemical energy storage power stations have the following characteristics: 1) Multiple safety assessment indicators must be established by integrating diverse information, including electrical, temperature, and gas information. 2) The safety status of battery cells and modules within the battery cabin must be assessed step by step, forming a multi-level safety assessment system that requires a high level of computational complexity. 3) The diverse and time-varying assessment factors demand rapid assessment. Therefore, it is necessary to conduct online safety assessment of batteries in large-capacity electrochemical energy storage power stations. However, there is currently a lack of safety assessment of electrochemical energy storage power stations during operation.

[0004] The key to online safety assessment of batteries in large-capacity electrochemical energy storage power stations is to build a reasonable evaluation index system and evaluation method. The safety assessment methods of energy storage systems are divided into two categories: model-driven and data-driven methods. The model-driven method is to conduct a comprehensive evaluation of multiple evaluation index values ​​obtained. For scenarios with a large number of evaluation objects and many evaluation indicators, the amount of calculation is large and the speed is limited. The data-driven method uses a machine learning algorithm to mine the nonlinear mapping relationship between the system safety score and multiple safety evaluation indicators. Its accuracy mainly depends on the generalization ability of the algorithm and the rationality of the training samples. Least Squares Support Vector Machine (LSSVM) is a machine learning algorithm with strong nonlinear mapping capabilities and the advantage of being suitable for small sample problems. Therefore, the present invention proposes a data-driven online safety assessment method and device for electrochemical energy storage power stations. Summary of the Invention

[0005] In order to solve the above problems, the present invention proposes a data-driven online safety assessment method and device for electrochemical energy storage power stations, which can solve the problem that the safety score data of training samples is difficult to obtain through measurement.

[0006] The technical solution adopted by the present invention to solve its technical problems is:

[0007] In a first aspect, an embodiment of the present invention provides a data-driven online safety assessment method for an electrochemical energy storage power station, comprising the following steps:

[0008] Step S1: Based on the structural characteristics of large-capacity electrochemical energy storage power stations and the current status of safety assessment development, a multi-level safety assessment index system for electrochemical energy storage power station batteries is constructed with multi-information fusion. The multi-level safety assessment index system for electrochemical energy storage power station batteries includes one target layer and three factor layers. The target layer is the online safety assessment result of the battery compartment of the energy storage power station. The first factor layer includes the battery module safety score, H2 volume fraction, and CO volume fraction indicators. The second factor layer includes the battery cell safety score, battery module voltage range, temperature range, voltage consistency, temperature consistency, and SOC consistency indicators. The third factor layer includes the battery cell voltage deviation rate, temperature, SOC, and SOH indicators.

[0009] Step S2, collecting historical monitoring data of a large-capacity electrochemical energy storage power station and calculating safety assessment index values ​​of a multi-level safety assessment index system for batteries in the electrochemical energy storage power station;

[0010] Step S3: Using the safety assessment index values ​​of the multi-level safety assessment index system for electrochemical energy storage power station batteries as input data samples, a model-driven battery safety assessment method based on game theory combined weighting-TOPSIS is used to obtain output data samples consisting of safety scores for battery cells, battery modules, and battery compartments, forming training sample data.

[0011] Step S4, using the training sample data to perform LSSVM network training to obtain a battery safety assessment LSSVM model and its parameters;

[0012] Step S5: obtaining the real-time safety assessment index value of the large-capacity electrochemical energy storage power station as the input of the battery safety assessment LSSVM model to obtain the battery cell safety score, battery module safety score and battery compartment online safety assessment results.

[0013] As a possible implementation of this embodiment, in step S1, the battery module safety score is obtained by evaluating the six indicators in the second factor layer, and the battery cell safety score is obtained by evaluating the four indicators in the third factor layer.

[0014] As a possible implementation of this embodiment, in step S1,

[0015] The battery cell safety score is the average of the scores of the battery cells in the same battery module;

[0016] The battery module safety score is the average of the scores of the battery modules in the same battery compartment.

[0017] As a possible implementation of this embodiment, in step S1, the battery module voltage extreme difference and temperature extreme difference indicators are the maximum voltage and temperature differences of different battery cells in the same battery module;

[0018] The battery cell voltage deviation rate, battery module voltage consistency, temperature consistency and SOC consistency indicators are defined as:

[0019]

[0020] Among them, △U battery cell voltage deviation rate index value, D VLR,x Where x is voltage, temperature, and SOC, which are the voltage consistency, temperature consistency, and SOC consistency index values ​​of the battery module respectively. n is the number of battery cells in the same battery module. U t is the battery cell voltage value at time t, is the average voltage of the battery cells in the same battery module at time t, x i and x are the voltage, temperature, SOC value of battery cell i in the same battery module at time t and their average values ​​in the same battery module.

[0021] As a possible implementation of this embodiment, step S2 includes:

[0022] The online monitoring BMS and battery compartment gas detectors can obtain the battery cell temperature, SOC and SOH index values, as well as the original values ​​of H2 volume fraction and CO volume fraction;

[0023] Based on the battery cell temperature, SOC and SOH data, the battery cell voltage deviation rate, battery module voltage extreme difference, temperature extreme difference and battery module voltage consistency, temperature consistency and SOC consistency index values ​​are calculated.

[0024] As a possible implementation of this embodiment, in step S3, the model-driven battery safety assessment method based on game theory combined weighting-TOPSIS is used to obtain output data samples consisting of safety scores of battery cells, battery modules, and battery compartments, including:

[0025] The original data matrix of the evaluation indicators is forward processed and then standardized to obtain a standardized matrix. Forward processing converts the safety evaluation indicators into extremely large indicators. In the multi-level safety evaluation indicator system, the SOH indicator is an extremely large indicator, the battery cell voltage offset rate, battery module voltage range, temperature range, battery module voltage consistency, temperature consistency, SOC consistency indicator, and H2 and CO volume fraction indicators are extremely small indicators, and the battery cell temperature and SOC indicators are interval indicators. The standardized matrix of n battery evaluation objects and m safety evaluation indicators is expressed as:

[0026]

[0027] Among them, Z is the normalization matrix, z 11 ,z 12 ,…,z 1m is the standard value of the 1st, 2nd, ..., mth safety assessment index value of the first battery assessment object after processing, z n1 ,z n2 ,…,z nm The processed standard values ​​of the 1st, 2nd, ..., mth safety assessment index values ​​of the nth battery assessment object;

[0028] Find the ideal optimal solution and the worst solution for each safety evaluation indicator:

[0029]

[0030] in, are the ideal optimal solution and the worst solution of the 1st, 2nd, ..., mth safety assessment indicators respectively, and max() and min() are the maximum and minimum functions respectively;

[0031] Calculate the distance between the battery evaluation object and the ideal optimal solution and the worst solution, and calculate their closeness. The closeness value is the battery safety score. The distance between the i-th battery evaluation object and the ideal optimal solution and the worst solution is:

[0032]

[0033] in, are the distances between the i-th battery evaluation object and the ideal optimal solution and the worst solution, respectively. are the ideal optimal solution and the worst solution of the j-th security evaluation index, respectively, z ij is the standard value of the jth safety assessment index value of the i-th battery assessment object after processing, ω j is the weight of the j-th security assessment indicator;

[0034] Calculate the closeness of the i-th battery evaluation object:

[0035]

[0036] Among them, E i is the closeness value of the i-th battery evaluation object, that is, the safety score value, 0≤E i ≤1, and E i The larger it is, the closer the battery evaluation object is to the ideal optimal solution, the higher the safety score value is, and the better the battery safety is.

[0037] As a possible implementation of this embodiment, the weights of the security assessment indicators are obtained using a game theory combined weighting method, including:

[0038] First, the AHP method and CRITIC method are used to obtain the subjective and objective weights of the safety assessment indicators. Then, game theory is used to balance the subjective and objective weights to obtain the comprehensive weights of the safety assessment indicators:

[0039] ω=α1ω1+α2ω2 (6)

[0040] Where ω is the linear combination weight of the evaluation indicators, ω1 and ω2 are the sets of subjective and objective weights of the evaluation indicators obtained by AHP and CRITIC, α1 and α2 are the proportions of subjective and objective weights, and α1+α2=1;

[0041] In order to minimize the deviation between the linear combination weights and the weighting results of AHP and CRITIC methods, we construct:

[0042]

[0043] According to the differential properties of the matrix, the optimal first-order derivative of formula (7) is calculated to obtain the linear equation system:

[0044]

[0045] Solve the linear equation system (8) to obtain α1, α2, and then obtain the comprehensive weight of the evaluation index:

[0046]

[0047] Among them, ω * is the comprehensive weight of the evaluation indicators.

[0048] In a second aspect, an embodiment of the present invention provides a data-driven online safety assessment device for an electrochemical energy storage power station, comprising:

[0049] An evaluation index system construction module is used to construct a multi-level safety evaluation index system for electrochemical energy storage power station batteries based on the structural characteristics of large-capacity electrochemical energy storage power stations and the current status of safety assessment development. The multi-level safety evaluation index system for electrochemical energy storage power station batteries includes one target layer and three factor layers. The target layer is the online safety evaluation result of the battery compartment of the energy storage power station. The first factor layer includes the battery module safety score, H2 volume fraction, and CO volume fraction indicators. The second factor layer includes the battery cell safety score, battery module voltage range, temperature range, voltage consistency, temperature consistency, and SOC consistency indicators. The third factor layer includes the battery cell voltage offset rate, temperature, SOC, and SOH indicators.

[0050] An evaluation index value calculation module is used to collect historical monitoring data of large-capacity electrochemical energy storage power stations and calculate the safety evaluation index values ​​of the multi-level safety evaluation index system of the electrochemical energy storage power station batteries;

[0051] The training sample data construction module is used to use the safety assessment index values ​​of the multi-level safety assessment index system of electrochemical energy storage power station batteries as input data samples, and adopt the model-driven battery safety assessment method based on game theory combined empowerment-TOPSIS to obtain output data samples consisting of battery cell, battery module and battery compartment safety scores to form training sample data;

[0052] LSSVM network training module, used to train the LSSVM network using training sample data to obtain the battery safety assessment LSSVM model and its parameters;

[0053] The online safety assessment module is used to obtain the real-time safety assessment index values ​​of large-capacity electrochemical energy storage power stations as input to the battery safety assessment LSSVM model to obtain the battery cell safety score, battery module safety score and battery compartment online safety assessment results.

[0054] In a third aspect, an embodiment of the present invention provides a computer device comprising a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform any of the steps of the above-mentioned data-driven online safety assessment method for electrochemical energy storage power stations.

[0055] In a fourth aspect, an embodiment of the present invention provides a storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of any of the above-mentioned data-driven online safety assessment methods for electrochemical energy storage power stations are executed.

[0056] The beneficial effects of the technical solutions of the embodiments of the present invention are as follows:

[0057] The technical solution of the embodiment of the present invention is a data-driven online safety assessment method for electrochemical energy storage power stations, which is both accurate and fast, and solves the problem that safety score data of training samples is difficult to obtain through measurement. It is suitable for safety assessment and health monitoring and early warning of batteries and systems of large-capacity electrochemical energy storage systems.

[0058] This paper proposes a model-driven battery safety assessment method based on game theory combined weighting - TOPSIS. It combines the TOPSIS comprehensive evaluation method with the game theory combined weighting based on the AHP method and the CRITIC method. It takes into account the comparison strength and conflict between evaluation indicators, as well as the subjectivity and objectivity of indicator weights. It is suitable for the mechanism modeling and analysis of electrochemical energy storage power station battery and system safety assessment under multiple influencing factors.

[0059] The present invention proposes a data-driven battery safety assessment method based on LSSVM. The method uses the results of a mechanism model to obtain training samples, and adopts the LSSVM regression model to extract battery safety assessment knowledge offline. The battery safety score is then obtained by LSSVM online evaluation. This method combines the accuracy and speed of safety assessment, and solves the problem that the safety score data of training samples is difficult to obtain through measurement. The method is suitable for safety assessment and health monitoring and early warning of batteries and systems of large-capacity electrochemical energy storage systems.

[0060] The data-driven online safety assessment device for an electrochemical energy storage power station according to the technical solution of an embodiment of the present invention has the same beneficial effects as the data-driven online safety assessment method for an electrochemical energy storage power station according to the technical solution of an embodiment of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a flow chart of a data-driven online safety assessment method for an electrochemical energy storage power station according to an exemplary embodiment;

[0062] Figure 2 1 is a schematic structural diagram of a data-driven online safety assessment device for an electrochemical energy storage power station according to an exemplary embodiment;

[0063] Figure 3 is a schematic diagram showing a battery cell safety scoring result according to an exemplary embodiment;

[0064] Figure 4 The figure is a schematic diagram showing a battery module safety scoring result according to an exemplary embodiment. DETAILED DESCRIPTION

[0065] In order to more clearly illustrate the technical features of the present invention, the present invention is described in detail below through specific implementation methods and in conjunction with the accompanying drawings.

[0066] like Figure 1 As shown, an embodiment of the present invention provides a data-driven online safety assessment method for an electrochemical energy storage power station, comprising the following steps:

[0067] Step S1: Based on the structural characteristics of large-capacity electrochemical energy storage power stations and the current status of safety assessment development, a multi-level safety assessment index system for electrochemical energy storage power station batteries is constructed with multi-information fusion. The multi-level safety assessment index system for electrochemical energy storage power station batteries includes one target layer and three factor layers. The target layer is the online safety assessment result of the battery compartment of the energy storage power station. The first factor layer includes the battery module safety score, H2 volume fraction, and CO volume fraction indicators. The second factor layer includes the battery cell safety score, battery module voltage range, temperature range, voltage consistency, temperature consistency, and SOC consistency indicators. The third factor layer includes the battery cell voltage deviation rate, temperature, SOC, and SOH indicators.

[0068] Step S2, collecting historical monitoring data of a large-capacity electrochemical energy storage power station and calculating safety assessment index values ​​of a multi-level safety assessment index system for batteries in the electrochemical energy storage power station;

[0069] Step S3: Using the safety assessment index values ​​of the multi-level safety assessment index system for electrochemical energy storage power station batteries as input data samples, a model-driven battery safety assessment method based on game theory combined weighting-TOPSIS is used to obtain output data samples consisting of safety scores for battery cells, battery modules, and battery compartments, forming training sample data.

[0070] Step S4, using the training sample data to perform LSSVM network training to obtain a battery safety assessment LSSVM model and its parameters;

[0071] In step S5, the real-time safety assessment index value of the large-capacity electrochemical energy storage power station is obtained as the input of the battery safety assessment LSSVM model, and the battery cell safety score, battery module safety score and battery compartment online safety assessment results are obtained, thus realizing the online safety assessment of the large-capacity electrochemical energy storage power station.

[0072] As a possible implementation of this embodiment, in step S1, the battery module safety score is obtained by evaluating the six indicators in the second factor layer, and the battery cell safety score is obtained by evaluating the four indicators in the third factor layer.

[0073] As a possible implementation of this embodiment, in step S1,

[0074] The battery cell safety score is the average of the scores of the battery cells in the same battery module;

[0075] The battery module safety score is the average of the scores of the battery modules in the same battery compartment.

[0076] As a possible implementation of this embodiment, in step S1, the battery module voltage extreme difference and temperature extreme difference indicators are the maximum voltage and temperature differences of different battery cells in the same battery module;

[0077] The battery cell voltage deviation rate, battery module voltage consistency, temperature consistency and SOC consistency indicators are defined as:

[0078]

[0079] Among them, △U battery cell voltage deviation rate index value, D VLR,x Where x is voltage, temperature, and SOC, which are the voltage consistency, temperature consistency, and SOC consistency index values ​​of the battery module respectively. n is the number of battery cells in the same battery module. U t is the battery cell voltage value at time t, is the average voltage of the battery cells in the same battery module at time t, x i 、 They are the voltage, temperature, SOC value of battery cell i in the same battery module at time t and their average values ​​in the same battery module.

[0080] As a possible implementation of this embodiment, step S2 includes:

[0081] The online monitoring BMS and battery compartment gas detectors can obtain the battery cell temperature, SOC and SOH index values, as well as the original values ​​of H2 volume fraction and CO volume fraction;

[0082] Based on the battery cell temperature, SOC and SOH data, the battery cell voltage deviation rate, battery module voltage extreme difference, temperature extreme difference and battery module voltage consistency, temperature consistency and SOC consistency index values ​​are calculated.

[0083] As a possible implementation of this embodiment, in step S3, the model-driven battery safety assessment method based on game theory combined weighting-TOPSIS is used to obtain output data samples consisting of safety scores of battery cells, battery modules, and battery compartments, including:

[0084] 1) The original data matrix of the evaluation indicators is forward processed and then standardized to obtain a standardized matrix; the forward processing is to convert the safety evaluation indicators into extremely large indicators. In the multi-level safety evaluation indicator system, the SOH indicator is an extremely large indicator, the battery cell voltage offset rate, battery module voltage range, temperature range, battery module voltage consistency, temperature consistency, SOC consistency indicator and H2 and CO volume fraction indicators are extremely small indicators, and the battery cell temperature and SOC indicators are interval indicators; the standardized matrix of n battery evaluation objects and m safety evaluation indicators is expressed as:

[0085]

[0086] Among them, Z is the normalization matrix, z 11 ,z 12 ,…,z 1m is the standard value of the 1st, 2nd, ..., mth safety assessment index value of the first battery assessment object after processing, z n1 ,z n2 ,…,z nm The processed standard values ​​of the 1st, 2nd, ..., mth safety assessment index values ​​of the nth battery assessment object;

[0087] 2) Find the ideal optimal solution and the worst solution for each safety assessment indicator:

[0088]

[0089] in, are the ideal optimal solution and the worst solution of the 1st, 2nd, ..., mth safety assessment indicators respectively, and max() and min() are the maximum and minimum functions respectively;

[0090] 3) Calculate the distance between the battery evaluation object and the ideal optimal solution and the worst solution, and calculate their closeness. The closeness value is the battery safety score. The distance between the i-th battery evaluation object and the ideal optimal solution and the worst solution is:

[0091]

[0092] in, are the distances between the i-th battery evaluation object and the ideal optimal solution and the worst solution, respectively. are the ideal optimal solution and the worst solution of the j-th security evaluation index, respectively, z ij is the standard value of the jth safety assessment index value of the i-th battery assessment object after processing, ω j is the weight of the j-th security assessment indicator;

[0093] Calculate the closeness of the i-th battery evaluation object:

[0094]

[0095] Among them, E i is the closeness value of the i-th battery evaluation object, that is, the safety score value, 0≤E i ≤1, and E i The larger it is, the closer the battery evaluation object is to the ideal optimal solution, the higher the safety score value is, and the better the battery safety is.

[0096] As a possible implementation of this embodiment, the weights of the security assessment indicators are obtained using a game theory combined weighting method, including:

[0097] First, the AHP method and CRITIC method are used to obtain the subjective and objective weights of the safety assessment indicators. Then, game theory is used to balance the subjective and objective weights to obtain the comprehensive weights of the safety assessment indicators:

[0098] ω=α1ω1+α2ω2 (6)

[0099] Where ω is the linear combination weight of the evaluation indicators, ω1 and ω2 are the sets of subjective and objective weights of the evaluation indicators obtained by AHP and CRITIC, α1 and α2 are the proportions of subjective and objective weights, and α1+α2=1;

[0100] In order to minimize the deviation between the linear combination weights and the weighting results of AHP and CRITIC methods, we construct:

[0101]

[0102] According to the differential properties of the matrix, the optimal first-order derivative of formula (7) is calculated to obtain the linear equation system:

[0103]

[0104] Solve the linear equations (8) to obtain α1, α2, and then obtain the comprehensive weight of the evaluation index:

[0105]

[0106] Among them, ω* is the comprehensive weight of the evaluation indicators.

[0107] like Figure 2 As shown, an embodiment of the present invention provides a data-driven online safety assessment device for an electrochemical energy storage power station, comprising:

[0108] An evaluation index system construction module is used to construct a multi-level safety evaluation index system for electrochemical energy storage power station batteries based on the structural characteristics of large-capacity electrochemical energy storage power stations and the current status of safety assessment development. The multi-level safety evaluation index system for electrochemical energy storage power station batteries includes one target layer and three factor layers. The target layer is the online safety evaluation result of the battery compartment of the energy storage power station. The first factor layer includes the battery module safety score, H2 volume fraction, and CO volume fraction indicators. The second factor layer includes the battery cell safety score, battery module voltage range, temperature range, voltage consistency, temperature consistency, and SOC consistency indicators. The third factor layer includes the battery cell voltage offset rate, temperature, SOC, and SOH indicators.

[0109] An evaluation index value calculation module is used to collect historical monitoring data of large-capacity electrochemical energy storage power stations and calculate the safety evaluation index values ​​of the multi-level safety evaluation index system of the electrochemical energy storage power station batteries;

[0110] The training sample data construction module is used to use the safety assessment index values ​​of the multi-level safety assessment index system of electrochemical energy storage power station batteries as input data samples, and adopt the model-driven battery safety assessment method based on game theory combined empowerment-TOPSIS to obtain output data samples consisting of battery cell, battery module and battery compartment safety scores to form training sample data;

[0111] LSSVM network training module, used to train the LSSVM network using training sample data to obtain the battery safety assessment LSSVM model and its parameters;

[0112] The online safety assessment module is used to obtain the real-time safety assessment index values ​​of large-capacity electrochemical energy storage power stations as input to the battery safety assessment LSSVM model to obtain the battery cell safety score, battery module safety score and battery compartment online safety assessment results.

[0113] The present invention first uses the evaluation index values ​​calculated from historical monitoring data as input data samples, and then adopts a model-driven battery safety assessment method based on game theory combined weighting-TOPSIS for the input data samples to obtain output data samples consisting of battery cell, battery module and battery compartment safety scores. The input data samples and output data samples are then used to form training sample data, and LSSVM training is performed to obtain a battery safety assessment LSSVM model and its parameters.

[0114] The safety assessment index value of the battery of the electrochemical energy storage power station to be evaluated at the current time t is then used as the input of the trained battery safety assessment LSSVM model. The safety scores of the battery cells, battery modules and battery compartments output by the LSSVM model are obtained in real time, realizing online safety assessment of large-capacity electrochemical energy storage power stations.

[0115] The batteries of an electrochemical energy storage power station are taken as the safety assessment objects. The battery type is lithium iron phosphate battery. Three battery compartments are selected, numbered 1 to 3. One battery compartment contains 24 modules, numbered 1 to 24. One module contains 6 battery cells. The data-driven online safety assessment method for large-capacity electrochemical energy storage power stations based on LSSVM proposed in this paper is used to conduct safety assessment.

[0116] 1) The comprehensive weight results of the safety assessment indicators are shown in Table 1.

[0117] Table 1: Weights of battery safety assessment indicators

[0118]

[0119] 2) The model-driven battery safety assessment method based on game theory combined weighting-TOPSIS proposed in this invention is used to obtain the battery safety score results. Among them, the safety scores of battery compartments 1 to 3 under a set of input data samples are 0.8199, 0.8562 and 0.8007. The input data samples of battery modules 1 to 5 in battery compartment 1 in this set of input data samples are shown in Table 2. The safety scores of battery modules 1 to 5 are 0.8467, 0.8712, 0.8225, 0.7638 and 0.7948, and the safety score results of the battery cells are shown in Table 2. Figure 3 shown.

[0120] Table 2: Input data samples for battery modules 1 to 5

[0121]

[0122] Combined with Table 2, Figure 3 The battery module safety scores show that the battery module safety score is determined by the values ​​of the battery module safety assessment indicators and their combined weights. The safety scores of the five modules and their battery cells range from 0.7 to 1, indicating good safety performance. Module 4 has a lower safety score than module 5 because the performance of each indicator in module 4, as shown in Table 2, is inferior to that of module 5. This verifies the effectiveness of the model-driven battery safety assessment method proposed in this paper based on game theory combined weighting - TOPSIS.

[0123] 3) At 30-second intervals, the real-time safety assessment index value data is input into the trained battery safety assessment LSSVM model for online safety assessment. For 100 sets of input data, the average relative errors of the safety scores of battery cells, battery modules, and battery compartments based on the model-driven safety score results are 4.2581%, 5.7197%, and 7.8896%, respectively. The safety score fitting curve of battery module 1 in battery compartment 1 under 30 sets of input data is shown in the figure below. Figure 4 As shown in Figure 2, the average time used for security assessment of a single set of input data under the model-driven and LSSVM models is 2.5697S and 0.0679S, respectively.

[0124] Depend on Figure 4 It can be seen from the average relative error of safety assessment and the evaluation time used that the average relative error of battery safety scores obtained by the data-driven battery safety assessment method based on LSSVM and the model-driven battery safety assessment method based on game theory combined empowerment-TOPSIS is less than 8%, the battery module safety score curve has a high degree of fitting, and the evaluation time is greatly shortened, which verifies the effectiveness and superiority of the data-driven large-capacity electrochemical energy storage power station online safety assessment method based on LSSVM proposed in the present invention.

[0125] A computer device provided in an embodiment of the present invention includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the device is running, the processor and the memory communicate via the bus, and the processor executes the machine-readable instructions to perform any of the steps of the above-mentioned data-driven electrochemical energy storage power station online safety assessment method.

[0126] Specifically, the above-mentioned memory and processor can be general-purpose memory and processor, which are not specifically limited here. When the processor runs the computer program stored in the memory, it can execute the above-mentioned data-driven electrochemical energy storage power station online safety assessment method.

[0127] Corresponding to the method for starting the above-mentioned application, an embodiment of the present invention further provides a storage medium on which a computer program is stored. When the computer program is run by a processor, the steps of any of the above-mentioned data-driven electrochemical energy storage power station online safety assessment methods are executed.

[0128] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A data-driven online safety assessment method for an electrochemical energy storage power station, characterized in that: The steps include: Step S1, based on the structural characteristics of large-capacity electrochemical energy storage power stations and the current status of safety assessment development, a multi-information fusion electrochemical energy storage power station battery multi-level safety assessment index system is constructed, wherein the electrochemical energy storage power station battery multi-level safety assessment index system includes 1 target layer and 3 factor layers, wherein the target layer is the online safety assessment result of the battery compartment of the energy storage power station, the first factor layer includes the battery module safety score, H2 volume fraction and CO volume fraction indicators, the second factor layer includes the battery cell safety score, battery module voltage extreme difference, temperature extreme difference, voltage consistency, temperature consistency and SOC consistency indicators, and the third factor layer includes the battery cell voltage deviation rate, temperature, SOC and SOH indicators; Step S2, collecting historical monitoring data of a large-capacity electrochemical energy storage power station, and calculating the safety assessment index value of a multi-level safety assessment index system of batteries in the electrochemical energy storage power station; Step S3, using the safety assessment index value of the multi-level safety assessment index system of the electrochemical energy storage power station battery as an input data sample, and using a model-driven battery safety assessment method based on game theory combined weighting-TOPSIS to obtain an output data sample consisting of safety scores of battery cells, battery modules and battery compartments, forming training sample data; Step S4, using the training sample data to perform LSSVM network training to obtain a battery safety assessment LSSVM model and its parameters; Step S5, obtaining the real-time safety assessment index value of the large-capacity electrochemical energy storage power station as the input of the battery safety assessment LSSVM model, and obtaining the battery cell safety score, battery module safety score and battery compartment online safety assessment results.

2. The data-driven online safety assessment method for electrochemical energy storage power station according to claim 1 is characterized in that: In step S1, the battery module safety score is obtained by evaluating the six indicators in the second factor layer, and the battery cell safety score is obtained by evaluating the four indicators in the third factor layer.

3. The data-driven online safety assessment method for electrochemical energy storage power station according to claim 1 is characterized in that: In step S1, The battery cell safety score is the average of the battery cell scores in the same battery module; The battery module safety score is the average of the scores of the battery modules in the same battery compartment.

4. The data-driven online safety assessment method for electrochemical energy storage power station according to claim 1 is characterized in that: In step S1, the battery module voltage extreme difference and temperature extreme difference indicators are the maximum voltage and temperature differences of different battery cells in the same battery module respectively; The battery cell voltage deviation rate, battery module voltage consistency, temperature consistency and SOC consistency indicators are defined as: Among them, △U is the battery cell voltage deviation rate index value, D VLR,x Where x is voltage, temperature, and SOC, which are the voltage consistency, temperature consistency, and SOC consistency index values ​​of the battery module, respectively; n is the number of battery cells in the same battery module; U t is the battery cell voltage value at time t, is the average voltage of the battery cells in the same battery module at time t, x i , They are respectively the voltage, temperature, SOC value of battery cell i in the same battery module at time t and their average values ​​in the same battery module.

5. The data-driven online safety assessment method for electrochemical energy storage power station according to claim 1 is characterized in that: The step S2 comprises: The battery cell temperature, SOC and SOH index values, as well as the original values ​​of H2 volume fraction and CO volume fraction are obtained from the online monitoring BMS and battery compartment gas detector; Based on the battery cell temperature, SOC and SOH data, the battery cell voltage deviation rate, battery module voltage extreme difference, temperature extreme difference and battery module voltage consistency, temperature consistency and SOC consistency index values ​​are calculated.

6. The data-driven online safety assessment method for an electrochemical energy storage power station according to any one of claims 1 to 5, characterized in that: In step S3, the model-driven battery safety assessment method based on game theory combined weighting-TOPSIS is used to obtain output data samples consisting of safety scores of battery cells, battery modules and battery compartments, including: The original data matrix of the evaluation index is forward processed and then standardized to obtain a standardized matrix; the forward processing is to convert the safety evaluation index into an extremely large index. The SOH index in the multi-level safety evaluation index system is an extremely large index, the battery cell voltage deviation rate, battery module voltage extreme difference, temperature extreme difference, battery module voltage consistency, temperature consistency, SOC consistency index and H2 and CO volume fraction index are extremely small indicators, and the battery cell temperature and SOC index are interval indicators; the standardized matrix of n battery evaluation objects and m safety evaluation indicators is expressed as: Among them, Z is the normalization matrix, z 11 ,z 12 ,…,z 1m is the standard value of the 1st, 2nd, …, mth safety assessment index value of the 1st battery assessment object after processing, z n1 ,z n2 ,…,z nm The standard values ​​of the 1st, 2nd, ..., mth safety assessment index values ​​of the nth battery assessment object after processing; Find the ideal optimal solution and the worst solution for each safety evaluation indicator: in, are the ideal optimal solution and the worst solution of the 1st, 2nd, …, mth safety assessment indicators respectively, and max() and min() are the maximum and minimum functions respectively; Calculate the distance between the battery evaluation object and the ideal optimal solution and the worst solution, and calculate their closeness. The closeness value is the battery safety score. The distance between the i-th battery evaluation object and the ideal optimal solution and the worst solution is: in, are the distances between the i-th battery evaluation object and the ideal optimal solution and the worst solution, respectively. are the ideal optimal solution and the worst solution of the jth safety evaluation index, respectively, ij is the standard value of the jth safety assessment index value of the i-th battery assessment object after processing, ω j is the weight of the jth security assessment indicator; Calculate the closeness of the i-th battery evaluation object: Among them, E i is the closeness value of the i-th battery evaluation object, that is, the safety score, 0≤E i ≤1, and E i The larger it is, the closer the battery evaluation object is to the ideal optimal solution, the higher the safety score is, and the better the battery safety is.

7. The data-driven online safety assessment method for electrochemical energy storage power station according to claim 6 is characterized in that: The weights of the safety assessment indicators are obtained using a game theory combined weighting method, including: Firstly, the AHP method and CRITIC method are used to obtain the subjective and objective weights of the safety assessment indicators. Then, game theory is used to balance the subjective and objective weights to obtain the comprehensive weights of the safety assessment indicators: ω=α1ω1+α2ω2 (6) Among them, ω is the linear combination weight of the evaluation index, ω1 and ω2 are the sets of subjective and objective weights of the evaluation index obtained by AHP and CRITIC, α1 and α2 are the proportions of subjective and objective weights, and α1+α2=1; In order to minimize the deviation between the linear combination weights and the weighting results of AHP and CRITIC methods, we construct: According to the differential properties of the matrix, the optimal first-order derivative of equation (7) is calculated to obtain the linear equation system: Solve the linear equation system (8) to obtain α1, α2, and then obtain the comprehensive weight of the evaluation index: Among them, ω * is the comprehensive weight of the evaluation indicators.

8. A data-driven online safety assessment device for an electrochemical energy storage power station, characterized in that: include: The evaluation index system construction module constructs a multi-level safety evaluation index system for electrochemical energy storage power station batteries based on the structural characteristics of large-capacity electrochemical energy storage power stations and the current status of safety evaluation. The multi-level safety evaluation index system for electrochemical energy storage power station batteries includes one target layer and three factor layers. The target layer is the online safety evaluation result of the battery compartment of the energy storage power station. The first factor layer includes the battery module safety score, H2 volume fraction, and CO volume fraction indicators. The second factor layer includes the battery cell safety score, battery module voltage extreme difference, temperature extreme difference, voltage consistency, temperature consistency, and SOC consistency indicators. The third factor layer includes the battery cell voltage deviation rate, temperature, SOC, and SOH indicators. An evaluation index value calculation module is used to collect historical monitoring data of large-capacity electrochemical energy storage power stations and calculate the safety evaluation index values ​​of the multi-level safety evaluation index system of the electrochemical energy storage power station battery; The training sample data composition module is used to use the safety assessment index value of the multi-level safety assessment index system of the electrochemical energy storage power station battery as the input data sample, and adopt the model-driven battery safety assessment method based on game theory combined weighting-TOPSIS to obtain the output data sample composed of the safety scores of the battery cell, battery module and battery compartment to constitute the training sample data; LSSVM network training module, used to train LSSVM network using training sample data to obtain battery safety assessment LSSVM model and its parameters; The online safety assessment module is used to obtain the real-time safety assessment index value of the large-capacity electrochemical energy storage power station as the input of the battery safety assessment LSSVM model to obtain the battery cell safety score, battery module safety score and battery compartment online safety assessment results.

9. A computer device, characterized in that: It includes a processor, a memory and a bus, the memory stores machine-readable instructions executable by the processor, when the computer device is running, the processor and the memory communicate through the bus, and the processor executes the machine-readable instructions to perform the steps of the data-driven electrochemical energy storage power station online safety assessment method as described in any one of claims 1-7.

10. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by a processor, executes the steps of the data-driven online safety assessment method for an electrochemical energy storage power station as described in any one of claims 1 to 7.