A method and system for detecting the operating status of an energy storage battery

Abstract

The present invention discloses a method and system for detecting the operating status of energy storage batteries. The method includes obtaining real-time data of each battery in a battery energy storage system and the battery series-parallel relationship. Based on the series structure and the real-time data of each battery, an abnormality probability vector for each series structure is calculated; the voltage abnormality probability vector and the temperature abnormality probability vector for the same series structure are integrated to obtain a series abnormality probability matrix; based on the real-time current data of each battery string in the parallel structure, a current abnormality probability vector for each battery string in the parallel structure is calculated; the current abnormality probability vector of each battery string in the parallel structure is weighted to the series abnormality probability matrix to obtain an abnormal state evaluation matrix; based on the abnormal state evaluation matrix, the operating status of each battery is determined to obtain the abnormal location of each battery in the battery energy storage system. This embodiment realizes real-time abnormality detection and location of battery energy storage operating status, improving the accuracy of battery energy storage abnormality warning.

Description

Technical Field

[0001] The present invention relates to the field of battery detection, and in particular to a method and system for detecting the operating status of an energy storage battery. Background Art

[0002] Battery energy storage systems are now widely used in my country's power system, with installed capacity increasing steadily. However, in recent years, safety incidents involving battery energy storage systems have been numerous worldwide, resulting in adverse social and economic impacts. Once a safety issue occurs in battery energy storage, the complex electrochemical processes within it can easily cause the battery to catch fire or even explode, creating a significant risk. Therefore, the application safety of battery energy storage systems has long been a key constraint on their further development. To prevent battery safety issues, it is imperative for battery energy storage system developers to proactively establish intelligent operations and maintenance management, anomaly warnings and fault identification, multi-layered safety protection, and multi-level fire protection based on the system's massive data collection.

[0003] At present, the safety management of battery energy storage systems mainly relies on multi-level battery management systems, local monitoring platforms, and multi-level fire protection systems (battery box level, battery cluster level, and battery compartment level). The relevant technologies have provided basic guarantees for the operation of battery energy storage. However, there are still problems such as insufficient mining of massive battery operation data, insufficient estimation of battery potential risks, and inaccurate battery warning and positioning technologies. Especially in the detection and positioning of abnormal battery operation status, current engineering applications mostly rely on two methods: (1) the battery management system detects that the voltage difference and temperature difference of multiple batteries under its management are large, and issues warning information by setting thresholds based on expert experience; (2) the local monitoring system issues warning information by comparing the real-time data reported by the battery management system with the battery historical data or through simple judgment methods. Although the above two methods form a preliminary judgment of the battery operation status at the battery management layer and the local monitoring layer, they are still relatively rough and simple, and the warning and positioning accuracy is insufficient.

[0004] In theoretical research, there have been many studies on data anomaly detection technology and fault location technology, and they have been used in some public test sets or some electrical equipment, achieving relatively good results. However, they have not been applied to battery energy storage anomaly detection. The core of this problem lies in the lack of depth in massive data mining and accuracy in data analysis, which leads to problems with the real-time and accuracy of anomaly detection and positioning. There is a lack of comprehensive battery energy storage operating status anomaly detection and positioning methods, the accuracy of battery energy storage abnormal status warning is low, and there are safety hazards in the battery application process. Summary of the Invention

[0005] The present invention provides a method and system for detecting the operating status of an energy storage battery, which realizes real-time abnormality detection and positioning of the battery energy storage operating status, is conducive to identifying and locating the abnormal operating status of each battery in the battery energy storage system, and improves the accuracy of the battery energy storage abnormality status warning.

[0006] In order to solve the above technical problems, an embodiment of the present invention provides a method for detecting the operating status of an energy storage battery, comprising:

[0007] Acquire real-time data of each battery in the battery energy storage system and the battery series-parallel relationship, and calculate the abnormality probability vector under each series structure based on the series structure and the real-time data of each battery; wherein the real-time data includes real-time voltage data, real-time temperature data, and real-time current data, the abnormality probability vector includes a voltage abnormality probability vector and a temperature abnormality probability vector, and the battery series-parallel relationship includes a series structure and a parallel structure;

[0008] The voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure are fused to obtain the series anomaly probability matrix.

[0009] Calculate the current abnormality probability vector of each battery string in the parallel structure based on the real-time current data of each battery string in the parallel structure;

[0010] The current abnormal probability vector of each battery string in the parallel structure is assigned to the series abnormal probability matrix to obtain the abnormal state evaluation matrix;

[0011] According to the abnormal state evaluation matrix, the operating state of each battery is determined, and the operating state detection result of each battery in the battery energy storage system is obtained; wherein the operating state detection result includes no abnormality and abnormal location.

[0012] In implementing the embodiments of the present invention, the abnormality probability vectors of each battery in the same series connection are calculated based on the real-time voltage and temperature values ​​of the series-connected batteries. The abnormality probability vectors obtained based on the voltage and temperature values ​​are then integrated using evidence fusion. Simultaneously, the abnormality probability vectors of each battery string in the parallel structure are calculated based on the battery current values ​​of each battery string in the parallel structure. The abnormality probability vectors integrated within each battery string are weighted by the abnormality probability of each battery string. Ultimately, an abnormality state evaluation matrix for the series-parallel structure of the battery energy storage system is obtained, i.e., an abnormality probability score for each battery. Based on the abnormal state evaluation matrix, the abnormality detection results and location of each battery operating state are determined. Compared with existing battery energy storage operation status abnormality detection and positioning methods, the integration of the real-time voltage, current and temperature original information of all batteries (no features are extracted here) can ensure the real-time detection. The integration of the voltage and temperature information of each battery in the same battery string, and the integration of the current information of each series battery branch can effectively improve the accuracy of battery operation status abnormality detection and form an accurate score for each battery operation status abnormality, which is conducive to the identification and positioning of each battery operation status abnormality in the battery energy storage system, and then form an effective early warning, improve the safety level of the battery energy storage system operation, and promote the more extensive and reliable and safe application of battery energy storage systems in new power systems.

[0013] As a preferred solution, real-time data of each battery in the battery energy storage system and the battery series and parallel relationship are obtained, specifically:

[0014] The real-time data of each battery in the battery energy storage system is measured by sensors to obtain the real-time data of each battery; wherein the sensors include voltage sensors, temperature sensors and current sensors;

[0015] According to the integrated structure of the battery energy storage system, the battery string and parallel situation is determined, and according to the battery string and parallel situation, each battery string is numbered in a two-dimensional array to determine the number of each battery; wherein, the battery string and parallel situation includes the number of each battery in the battery string, the number of parallel battery strings, and the battery string and parallel relationship.

[0016] As a preferred solution, based on the series structure and the real-time data of each battery, the abnormal probability vector under each series structure is calculated, specifically:

[0017] Calculating a deviation matrix under the current series structure based on the absolute value of the difference between the real-time data of each battery in the current series structure and the real-time data of two adjacent batteries; wherein the deviation matrix includes a voltage deviation matrix and a current deviation matrix;

[0018] According to the deviation matrix under the current series structure, the deviation matrices under several series structures are obtained;

[0019] According to the deviation matrix under each series structure, the abnormal probability vector under each series structure is calculated.

[0020] As a preferred solution, the abnormal probability vector under each series structure is calculated according to the deviation matrix under each series structure, specifically:

[0021] Calculate the affinity matrix under the current series structure according to the deviation matrix under the current series structure and the preset similarity entropy value range;

[0022] Normalizing the affinity matrix under the current series structure to obtain a probability matrix under the current series structure; wherein the probability matrix includes a voltage probability matrix and a temperature probability matrix;

[0023] According to the probability matrix under the current series structure, the abnormal state scoring vector of each battery in the current series structure is calculated. The abnormal state scoring vector includes the voltage abnormal state scoring vector and the temperature abnormal state scoring vector. The formula is as follows:

[0024]

[0025] Among them, c U i,s(t) is the voltage abnormality score vector of the sth battery in the ith battery string at time t, b U i,j,s(t) is the voltage probability matrix value of the sth battery under the i-th battery string at time t;

[0026]

[0027] Among them, c Tem i,s(t) is the temperature abnormality status score vector of the sth battery under the ith battery string at time t, b Tem i,j,s(t) is the temperature probability matrix value of the sth battery under the i-th battery string at time t;

[0028] Normalize the abnormal state score vector of each battery in the current series structure to obtain the abnormal probability vector in the current series structure;

[0029] According to the abnormal probability vector under the current series structure, the abnormal probability vector under each series structure is obtained. The formula is as follows:

[0030] P U i(t)=[p U i,1 (t),p U i,2 (t),…,p U i,m (t)] T

[0031]

[0032] Among them, p U i,s(t) is the voltage abnormality probability vector of the sth battery under the ith battery string at time t, P U i(t) is the voltage abnormality probability vector of the i-th battery string at time t;

[0033] P Tem i(t)=[p Tem i,1(t),p Tem i,2(t),…,p Tem i,m(t)] T

[0034]

[0035] Among them, p Tem i,s(t) is the temperature anomaly probability vector of the sth battery under the ith battery string at time t, P Tem i(t) is the temperature anomaly probability vector of the i-th battery string at time t.

[0036] As a preferred solution, the affinity matrix under the current series structure is calculated according to the deviation matrix under the current series structure and the preset similarity entropy value range, specifically:

[0037] Obtaining a deviation value of each battery according to the deviation matrix under the current series structure; wherein the deviation value includes a voltage deviation value and a temperature deviation value;

[0038] According to the deviation value of each battery, the preset similarity entropy value range, the similarity entropy relationship and the similarity deviation relationship, the standard deviation of each battery is calculated by the binary search method. The similarity entropy relationship includes the voltage similarity entropy relationship and the temperature similarity entropy relationship. The voltage similarity entropy relationship is as follows:

[0039]

[0040] Among them, H(d U i,s(t)) is the voltage similarity entropy between the sth battery and the rth battery in the i-th battery string at time t, d U i,r|i,s (t) is the voltage similarity between the sth battery and the rth battery in the i-th battery string at time t;

[0041] The temperature similarity entropy relationship is as follows:

[0042]

[0043] Among them, H(d Tem i,s(t)) is the temperature similarity entropy between the sth battery and the rth battery in the i-th battery string at time t, dTem i,r|i,s (t) is the temperature similarity between the sth battery and the rth battery in the i-th battery string at time t;

[0044] Among them, the similarity deviation relationship includes the voltage similarity deviation relationship and the temperature similarity deviation relationship. The voltage similarity deviation relationship is as follows:

[0045]

[0046] Among them, σ U i,s (t) is the voltage standard deviation of the sth battery in the ith battery string at time t, ΔU i,s,r (t) is the voltage deviation between the sth battery and the rth battery in the i-th battery string at time t, ΔU i,s,k (t) is the voltage deviation between the sth battery and the kth battery in the i-th battery string at time t;

[0047] The temperature similarity deviation relationship is as follows:

[0048]

[0049] Among them, σ Tem i,s (t) is the temperature standard deviation of the sth battery in the ith battery string at time t, ΔT i,s,r (t) is the temperature deviation between the sth battery and the rth battery in the i-th battery string at time t, ΔT i,s,k (t) is the temperature deviation between the sth battery and the kth battery in the i-th battery string at time t;

[0050] Based on the similarity deviation relationship, the standard deviation and deviation value of each battery in the current series structure, the affinity matrix under the current series structure is calculated. The affinity matrix includes the voltage affinity matrix and the temperature affinity matrix. The formula is as follows:

[0051]

[0052] Among them, A U i(t) is the voltage affinity matrix of the i-th battery string at time t, is the voltage similarity between the jth battery and the mth battery in the i-th battery string at time t;

[0053]

[0054] Among them, A Tem i(t) is the temperature affinity matrix of the i-th battery string at time t, is the temperature similarity between the jth battery and the mth battery in the i-th battery string at time t.

[0055] As a preferred solution, the voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure are fused to obtain the series anomaly probability matrix, which is:

[0056] The voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure are fused to obtain the series anomaly probability matrix under the same series structure. The evidence fusion formula is as follows:

[0057] P S i(t)=[P S i,1(t),P S i,2(t),…,P S i,m(t)] T

[0058]

[0059] Among them, P S i(t) is the probability vector of abnormal connection of the i-th battery string at time t, The voltage anomaly probability vector and temperature anomaly probability vector of the i-th battery string at time t are fused together to form the B-th battery string. i,r The probability of battery abnormality, is the Bth battery string under the i-th battery string at time t i,r The voltage abnormality probability vector of each battery, is the Bth battery string under the i-th battery string at time t i,r The temperature anomaly probability vector of the battery, k c The conflict coefficient is equal to

[0060] According to the series anomaly probability matrix under the same series structure, the series anomaly probability matrix is ​​obtained, and the formula is as follows:

[0061] P S (t)=[P S 1(t),P S 1(t),…,P S n (t)]∈R m×n

[0062] Among them, P S (t) is the concatenated anomaly probability vector.

[0063] As a preferred solution, the current abnormality probability vector of each battery string in the parallel structure is weighted to the series abnormality probability matrix to obtain the abnormal state evaluation matrix, which is specifically:

[0064] The current abnormal probability vectors of each battery string in the parallel structure are assigned column-wise weights to the series abnormal probability matrix to obtain the abnormal state evaluation matrix, which is specifically:

[0065] S C (t)=[p I 1(t)×P1 S (t),p I 2(t)×P2 S (t),…,p I n (t)×P n S (t)]∈R m×n

[0066] Among them, S C (t) is the abnormal state evaluation matrix, P S n(t) is the probability vector of the series abnormality of the nth battery string at time t, p I n (t) The current abnormality probability vector of the nth battery string at time t.

[0067] As a preferred solution, the operating status of each battery is determined according to the abnormal status evaluation matrix, specifically:

[0068] Calculate the mean score based on the abnormal state evaluation matrix;

[0069] The operating status of each battery is determined based on the relationship between all elements in the abnormal status evaluation matrix and the mean score.

[0070] As a preferred solution, the operating status of each battery is determined based on the relationship between all elements in the abnormal status evaluation matrix and the mean score, specifically:

[0071] If the abnormal state evaluation matrix value of the element at position (i, j) is not greater than the mean score, the battery operation state of the i-th battery in the j-th series battery branch is determined to be normal;

[0072] If the abnormal state evaluation matrix value of the element at the (i, j)th position is greater than the mean score, the battery operation state of the i-th battery in the j-th series battery branch is determined to be abnormal.

[0073] In order to solve the same technical problem, an embodiment of the present invention further provides a detection system for the operating status of an energy storage battery, comprising: a series anomaly module, an evidence fusion module, a parallel anomaly module, a weighting module and a status evaluation module;

[0074] The series anomaly module is used to obtain the real-time data of each battery in the battery energy storage system and the series-parallel relationship of the batteries. Based on the series structure and the real-time data of each battery, the anomaly probability vector under each series structure is calculated. The real-time data includes real-time voltage data, real-time temperature data, and real-time current data. The anomaly probability vector includes a voltage anomaly probability vector and a temperature anomaly probability vector. The series-parallel relationship of the batteries includes a series structure and a parallel structure.

[0075] The evidence fusion module is used to fuse the voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure to obtain the series anomaly probability matrix;

[0076] The parallel abnormality module is used to calculate the current abnormality probability vector of each battery string in the parallel structure based on the real-time current data of each battery string in the parallel structure;

[0077] The weighting module is used to assign the current abnormality probability vector of each battery string in the parallel structure to the series abnormality probability matrix to obtain the abnormal state evaluation matrix;

[0078] The status evaluation module is used to determine the operating status of each battery according to the abnormal status evaluation matrix and obtain the operating status detection results of each battery in the battery energy storage system; wherein the operating status detection results include no abnormality and abnormal location. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 : A flow chart of an embodiment of a method for detecting the operating status of an energy storage battery provided by the present invention;

[0080] Figure 2 : A calculation flow chart of an embodiment of a method for detecting the operating status of an energy storage battery provided by the present invention;

[0081] Figure 3 : A schematic diagram of a battery energy storage system integration according to an embodiment of a method for detecting the operating status of an energy storage battery provided by the present invention;

[0082] Figure 4 : A flow chart of voltage anomaly probability matrix calculation for an embodiment of a method for detecting the operating status of an energy storage battery provided by the present invention;

[0083] Figure 5 : A structural diagram of X of an embodiment of a detection system for the operating status of an energy storage battery provided by the present invention. DETAILED DESCRIPTION

[0084] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0085] Example 1

[0086] Please refer to Figure 1 , which is a flow chart of a method for detecting the operating status of an energy storage battery provided by an embodiment of the present invention, the calculation process is as follows Figure 2 The battery operating status detection method of this embodiment is applicable to the abnormal status detection of each battery in the battery energy storage system. This embodiment improves the accuracy of the battery energy storage abnormal status warning by real-time abnormality detection and positioning of the battery energy storage operating status. The battery operating status detection method includes steps 101 to 105, each of which is as follows:

[0087] Step 101: Acquire real-time data of each battery in the battery energy storage system and the battery series-parallel relationship. Calculate an abnormality probability vector for each series structure based on the series structure and the real-time data of each battery. The real-time data includes real-time voltage data, real-time temperature data, and real-time current data. The abnormality probability vector includes a voltage abnormality probability vector and a temperature abnormality probability vector. The battery series-parallel relationship includes a series structure and a parallel structure.

[0088] Optionally, step 101 specifically includes steps 1011 to 1012, and each step is specifically as follows:

[0089] Step 1011: Real-time data of each battery in the battery energy storage system is measured using sensors to obtain real-time data of each battery; wherein the sensors include voltage sensors, temperature sensors, and current sensors; the battery string and parallel connection status is determined based on the integrated structure of the battery energy storage system, and each battery string is numbered in a two-dimensional array based on the battery string and parallel connection status to determine the battery number; wherein the battery string and parallel connection status includes the number of each battery in the battery string, the number of parallel battery strings, and the battery string and parallel connection relationship.

[0090] In this embodiment, according to the battery energy storage system integration mechanism, the number of batteries m in each battery string, the number of parallel battery strings n, and the battery string-parallel relationship in the battery energy storage system are determined, and the battery energy storage system integration diagram is as follows: Figure 3 According to the battery energy storage system integration structure, the number of batteries m in each battery string and the number of battery strings in parallel n in the battery energy storage system are determined, and each string of batteries is numbered in a two-dimensional array according to the string and parallel situation. For example, the jth battery under the i-th battery string can be defined as B i,j,i=1,2,…,n,j=1,2,…,m,;the voltage, temperature and current measurements at time t are U i,j (t), T i,j (t) and I i,j (t), where all battery currents in the i-th battery string are the same and are defined as I i (t), that is, I i,s (t) = I i,r (t) = I i (t),

[0091] Step 1012: Calculate the deviation matrix under the current series structure based on the absolute value of the difference between the real-time data of each battery in the current series structure and the real-time data of two adjacent batteries; wherein the deviation matrix includes a voltage deviation matrix and a current deviation matrix; obtain the deviation matrices under several series structures based on the deviation matrix under the current series structure; calculate the abnormal probability vector under each series structure based on the deviation matrix under each series structure.

[0092] In this embodiment, the voltage anomaly probability matrix calculation process is as follows: Figure 4 As shown, the real-time voltage data of all batteries measured by voltage sensors are used to calculate the voltage deviation matrix ΔU between each battery in the same series structure, and then the voltage abnormality probability vector of each battery in the same series structure is calculated to form the temperature abnormality probability matrix P under all series branches. U Similar to the voltage calculation, the real-time temperature data of all batteries measured by temperature sensors are used to calculate the temperature deviation matrix ΔT between each battery in the same series structure, and then the temperature anomaly probability vector of each battery in the same series structure is calculated to form the temperature anomaly probability matrix P for all series branches. Tem .

[0093] In this embodiment, the voltage deviation matrix ΔU between the batteries in the same series structure is calculated based on the real-time voltage data of all batteries measured by the voltage sensor. The voltage deviation matrix of each battery in the i-th battery string at time t is ΔU i (t), the calculation formula is as follows:

[0094]

[0095] It should be noted that this matrix represents the voltage deviation value, U i,m represents the mth battery in the i-th battery string, |ΔU i,1 -ΔU i,2| represents the absolute value of the difference between the voltage of the first battery and the voltage of the second battery in the i-th battery string; calculating the voltage deviation matrix is ​​to find the absolute value of the voltage difference between each two of the m batteries in the i-th battery string, forming an m×m matrix, and R mathematically represents a set of real numbers.

[0096] In this embodiment, the temperature deviation matrix ΔT between the batteries in the same series structure is calculated based on the real-time temperature data of all batteries measured by the temperature sensor. The temperature deviation matrix of each battery in the i-th battery string at time t is ΔT i (t), the calculation formula is as follows:

[0097]

[0098] Optionally, according to the deviation matrix under each series structure, the abnormal probability vector under each series structure is calculated, specifically including steps 1 to 5, each step is as follows:

[0099] Step 1: Calculate the affinity matrix under the current serial structure according to the deviation matrix under the current serial structure and the preset similarity entropy value range.

[0100] Optionally, the first step specifically includes: obtaining a deviation value of each battery according to a deviation matrix under the current series structure; wherein the deviation value includes a voltage deviation value and a temperature deviation value;

[0101] According to the deviation value of each battery, the preset similarity entropy value range, the similarity entropy relationship and the similarity deviation relationship, the standard deviation of each battery is calculated by the binary search method. The similarity entropy relationship includes the voltage similarity entropy relationship and the temperature similarity entropy relationship. The voltage similarity entropy relationship is as follows:

[0102]

[0103] Among them, H(d U i,s(t)) is the voltage similarity entropy between the sth battery and the rth battery in the i-th battery string at time t, d U i,r|i,s (t) is the voltage similarity between the sth battery and the rth battery in the i-th battery string at time t;

[0104] The temperature similarity entropy relationship is as follows:

[0105]

[0106] Among them, H(d Tem i,s(t)) is the temperature similarity entropy between the sth battery and the rth battery in the i-th battery string at time t, d Tem i,r|i,s(t) is the temperature similarity between the sth battery and the rth battery in the i-th battery string at time t;

[0107] Among them, the similarity deviation relationship includes the voltage similarity deviation relationship and the temperature similarity deviation relationship. The voltage similarity deviation relationship is as follows:

[0108]

[0109] Among them, σ U i,s (t) is the voltage standard deviation of the sth battery in the ith battery string at time t, ΔU i,s,r (t) is the voltage deviation between the sth battery and the rth battery in the i-th battery string at time t, ΔU i,s,k (t) is the voltage deviation between the sth battery and the kth battery in the i-th battery string at time t;

[0110] The temperature similarity deviation relationship is as follows:

[0111]

[0112] Among them, σ Tem i,s (t) is the temperature standard deviation of the sth battery in the ith battery string at time t, ΔT i,s,r (t) is the temperature deviation between the sth battery and the rth battery in the i-th battery string at time t, ΔT i,s,k (t) is the temperature deviation between the sth battery and the kth battery in the i-th battery string at time t;

[0113] Based on the similarity deviation relationship, the standard deviation and deviation value of each battery in the current series structure, the affinity matrix under the current series structure is calculated. The affinity matrix includes the voltage affinity matrix and the temperature affinity matrix. The formula is as follows:

[0114]

[0115] Among them, A U i(t) is the voltage affinity matrix of the i-th battery string at time t, is the voltage similarity between the jth battery and the mth battery in the i-th battery string at time t;

[0116]

[0117] Among them, A Tem i(t) is the temperature affinity matrix of the i-th battery string at time t, is the temperature similarity between the jth battery and the mth battery in the i-th battery string at time t.

[0118] In this embodiment, when calculating the voltage standard deviation of each battery in each battery string according to the voltage deviation matrix, the standard deviation σ of the sth battery in the ith battery string at time t is defined as U i,s(t), and defines the voltage similarity d between the sth battery and the rth battery in the i-th battery string at time t U i,r|i,s(t), the voltage similarity deviation relationship is calculated as follows:

[0119]

[0120] The voltage similarity entropy value H(d U i,s(t)), the voltage similarity entropy relationship formula is as follows:

[0121]

[0122] Then, by reasonably setting the entropy value H(d U i,s(t)) range, (preferably 2.32 to 5.64), using binary search to calculate the voltage standard deviation σ of the sth battery in the i-th battery string at time t U i,s(t);

[0123] The voltage standard deviation, voltage deviation matrix and voltage similarity calculation formula (voltage similarity deviation relationship calculation formula) of each battery in each battery string are calculated to calculate the voltage affinity matrix A of each battery in each battery string. U , that is, the voltage affinity matrix A of the i-th battery string at time t i U (t), expressed as:

[0124]

[0125] in, The assignment of uniform symbolic variables is only for the convenience of explanation and has no special physical meaning;

[0126] In this embodiment, when calculating the temperature standard deviation of each battery in each battery string according to the temperature deviation matrix, the standard deviation σ of the sth battery in the i-th battery string at time t is defined as Tem i,s(t), and defines the temperature similarity d between the sth battery and the rth battery in the i-th battery string at time t Tem i,r|i,s(t), the temperature similarity deviation relationship is calculated as follows:

[0127]

[0128] The temperature similarity entropy value H(dTem i,s(t)), the temperature similarity entropy relationship formula is as follows:

[0129]

[0130] Then, by reasonably setting the entropy value H(d Tem i,s(t)) range, (preferably 2.32 to 5.64), using the binary search method to calculate the temperature standard deviation σ of the sth battery in the ith battery string at time t Tem i,s(t);

[0131] The calculated temperature standard deviation, temperature deviation matrix and temperature similarity calculation formula (temperature similarity deviation relationship calculation formula) of each battery in each battery string are used to calculate the temperature affinity matrix A of each battery in each battery string. Tem , that is, the temperature affinity matrix A of the i-th battery string at time t i Tem (t), expressed as:

[0132]

[0133] in, The assignment of uniform symbolic variables is only for the convenience of explanation and has no special physical meaning;

[0134] Step 2: Normalize the affinity matrix under the current series structure to obtain the probability matrix under the current series structure; the probability matrix includes a voltage probability matrix and a temperature probability matrix.

[0135] In this embodiment, the voltage affinity matrix in step 1 is normalized to obtain the voltage probability matrix B of each battery in each battery string. U That is, the voltage probability matrix B of the i-th battery string at time t is i U (t) can be expressed as:

[0136]

[0137] Normalize the temperature affinity matrix in step 1 to obtain the temperature probability matrix B of each battery in each battery string Tem That is, the temperature probability matrix B of the i-th battery string at time t is i Tem (t) can be expressed as:

[0138]

[0139] Step 3: Calculate the abnormal state score vector of each battery in the current series structure based on the probability matrix in the current series structure. The abnormal state score vector includes the voltage abnormal state score vector and the temperature abnormal state score vector. The formula is as follows:

[0140]

[0141] Among them, c U i,s(t) is the voltage abnormality score vector of the sth battery in the ith battery string at time t, b U i,j,s(t) is the voltage probability matrix value of the sth battery under the i-th battery string at time t;

[0142]

[0143] Among them, c Tem i,s(t) is the temperature abnormality status score vector of the sth battery under the ith battery string at time t, b Tem i,j,s(t) is the temperature probability matrix value of the sth battery under the i-th battery string at time t.

[0144] In this embodiment, the voltage abnormality state score vector of each battery in the i-th battery string at time t is C U i,s (t)=[c U i,1 (t),c U i,2 (t),…,c U i,m (t)] T At time t, the temperature abnormality score vector of each battery in the i-th battery string is C Tem i,s (t)=[c Tem i,1 (t),c Tem i,2 (t),…,c Tem i,m (t)] T .

[0145] Step 4: Normalize the abnormal state score vector of each battery in the current series structure to obtain the abnormal probability vector in the current series structure.

[0146] In this embodiment, the voltage abnormality state score vectors of each battery in the same series structure obtained in step 3 are normalized to obtain the voltage abnormality probability vector P U i(t), that is, P U i (t)=[p Ui,1 (t),p U i,2 (t),…,p U i,m (t)] T ,in,

[0147] In this embodiment, the temperature abnormality state score vectors of each battery in the same series structure obtained in step 3 are normalized to obtain the voltage abnormality probability vector P Tem i(t), that is, P Tem i (t)=[p Tem i,1 (t),p Tem i,2 (t),…,p Tem i,m (t)] T ,in, P I i(t)=P Tem i(t).

[0148] Step 5: According to the abnormal probability vector under the current series structure, obtain the abnormal probability vector under each series structure. The formula is as follows:

[0149] P U i(t)=[p U i,1 (t),p U i,2 (t),…,p U i,m (t)] T

[0150]

[0151] Among them, p U i,s(t) is the voltage abnormality probability vector of the sth battery under the ith battery string at time t, P U i(t) is the voltage abnormality probability vector of the i-th battery string at time t;

[0152] P Tem i(t)=[p Tem i,1(t),p Tem i,2(t),…,p Tem i,m(t)] T

[0153]

[0154] Among them, p Temi,s(t) is the temperature anomaly probability vector of the sth battery under the ith battery string at time t, P Tem i(t) is the temperature anomaly probability vector of the i-th battery string at time t.

[0155] In this embodiment, the voltage anomaly probability vector is continuously calculated according to steps 1 and 5 until the voltage anomaly matrix P of all n battery strings at time t is generated. U (t), that is, the voltage anomaly probability vector under each series structure, is expressed as follows:

[0156] P U (t)=[P U 1(t),P U 1(t),…,P U n (t)]∈R m×n

[0157] Continuously calculate the temperature anomaly probability vector according to steps 1 and 5 until the temperature anomaly matrix P of all n battery strings at time t is generated. Tem (t), that is, the temperature anomaly probability vector under each series structure, its expression is as follows: P Tem (t)=[P Tem 1(t),P Tem 1(t),…,P Tem n (t)]∈R m×n

[0158] Step 102: Perform evidence fusion on the voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure to obtain a series anomaly probability matrix.

[0159] Optionally, step 102 specifically includes: fusing the voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure to obtain a series anomaly probability matrix under the same series structure. The evidence fusion formula is as follows:

[0160] P S i(t)=[P S i,1(t),P S i,2(t),…,P S i,m(t)] T

[0161]

[0162] Among them, P S i(t) is the probability vector of abnormal connection of the i-th battery string at time t, The voltage anomaly probability vector and temperature anomaly probability vector of the i-th battery string at time t are fused together to form the B-th battery string. i,r The probability of battery abnormality, is the Bth battery string under the i-th battery string at time t i,r The voltage abnormality probability vector of each battery, is the Bth battery string under the i-th battery string at time t i,r The temperature anomaly probability vector of the battery, k c The conflict coefficient is equal to

[0163] According to the series anomaly probability matrix under the same series structure, the series anomaly probability matrix is ​​obtained, and the formula is as follows:

[0164] P S (t)=[P S 1(t),P S 1(t),…,P S n (t)]∈R m×n

[0165] Among them, P S (t) is the concatenated anomaly probability vector.

[0166] In this embodiment, the identification framework of the evidence theory of each battery in the same series structure is defined. The identification framework of the i-th battery string at time t is {B i,1 ,B i,2 ,…,B i,m}, the voltage and temperature anomaly probability vector of the battery string is P U i(t) and P Tem i(t);

[0167] The evidence fusion formula is used to calculate the voltage and temperature anomaly probability vectors of the batteries in the same series structure after fusion to form a series anomaly probability vector. That is, the voltage and temperature anomaly probability vectors of the i-th battery string at time t are calculated as follows:

[0168]

[0169] Among them, it means It represents the voltage and temperature anomaly probability vector evidence fusion of the ith battery string at time t. i,r The probability of battery abnormality, k c The conflict coefficient is equal to Then the abnormal probability vector P of the series connection of the i-th battery string at time t is S i (t)=[P S i,1 (t),P Si,2 (t),…,P S i,m (t)] T .

[0170] Continuously calculate until the voltage and temperature anomaly probability vectors of all n battery strings at time t are generated and the series anomaly probability matrix P is generated. S (t), which is expressed as follows:

[0171] P S (t)=[P S 1(t),P S 1(t),…,P S n (t)]∈R m×n

[0172] Step 103: Calculate the current abnormality probability vector of each battery string in the parallel structure based on the real-time current data of each battery string in the parallel structure.

[0173] In this embodiment, based on the real-time current data of each battery string in the parallel structure, the calculation process of the current anomaly probability vector of each battery string in the parallel structure is similar to the calculation of the voltage anomaly probability vector and the temperature anomaly probability vector in step 101. The real-time current data of each battery string branch measured by the current sensor is used to calculate the current deviation matrix ΔI between each battery string branch (parallel structure), and then the current anomaly probability vector P between each battery string branch is calculated. I ;

[0174] Based on the real-time current data of each battery string branch measured by the current sensor, the current deviation matrix ΔI between each battery string branch (parallel structure) is calculated as follows, where the current of the i-th battery string branch at time t is I i (t):

[0175]

[0176] According to the current deviation matrix ΔI, the current abnormal probability vector P between each battery string branch (parallel structure) is calculated I , where the probability vector of abnormal current in the i-th battery string branch at time t is P I i (t);

[0177] When calculating the current standard deviation of each battery string branch according to the current deviation matrix, the standard deviation σ of the i-th battery string branch at time t is defined as I i (t). At the same time, the current similarity d in the i-th battery string branch and the i-th battery string branch at time t is defined I j|i(t) is calculated as follows:

[0178]

[0179] Define the current similarity entropy value H(d I i (t)) is:

[0180]

[0181] Then, by reasonably setting the entropy value H(dI i(t)) range (recommended range 2.32~5.64), the current standard deviation σ of the i-th battery string branch at time t is calculated using the binary search method. I i (t);

[0182] The current standard deviation, current deviation matrix and current similarity calculation formula of each battery string branch are calculated, and the current affinity matrix A of each battery string branch is calculated. I , that is, the current affinity matrix A of each battery string branch at time t I (t) can be expressed as

[0183]

[0184] Normalize the current affinity matrix to obtain the current probability matrix B between each battery string I That is, the current probability matrix B at time t I (t) can be expressed as

[0185]

[0186] The calculated current probability matrix B between each battery string I , calculate the current abnormal state score vector of each battery string, that is, the current abnormal state score vector c of the i-th battery string at time t I i (t) can be expressed as:

[0187]

[0188] The current abnormal state score vector between each battery string branch at time t is:

[0189] C I i (t)=[c I 1(t),c I 2(t),…,c I n (t)]

[0190] Normalize the current abnormal state score vector of each battery string branch to obtain the current abnormal probability vector P I (t), that is, P I (t)=[pI 1(t),pI 2(t),…,pI n(t)] T ,in

[0191] Step 104: Assign the current abnormality probability vector of each battery string in the parallel structure to the series abnormality probability matrix to obtain an abnormal state evaluation matrix.

[0192] Optionally, step 104 specifically includes assigning column-wise weights to the series abnormality probability matrix for the current abnormality probability vectors of each battery string in the parallel structure to obtain an abnormal state evaluation matrix, specifically:

[0193] S C (t)=[p I 1(t)×P1 S (t),p I 2(t)×P2 S (t),…,p I n (t)×P n S (t)]∈R m×n

[0194] Among them, S C (t) is the abnormal state evaluation matrix, P S n(t) is the probability vector of the series abnormality of the nth battery string at time t, p I n (t) The current abnormality probability vector of the nth battery string at time t.

[0195] In this embodiment, the current abnormality probability vector P of each battery string branch is used. I (t) The series abnormality probability matrix P under column-by-column weighting of all series branches S (t), obtain the abnormal state evaluation matrix S of all m×n batteries c (t), calculated as follows: S C (t)=[p I 1(t)×P1 S (t),p I 2(t)×P2 S (t),…,p I n (t)×P n S (t)]∈R m×n

[0196] Step 105: Determine the operating status of each battery according to the abnormal status evaluation matrix, and obtain the operating status detection result of each battery in the battery energy storage system; wherein the operating status detection result includes no abnormality and abnormal location.

[0197] Optionally, the operating status of each battery is determined according to the abnormal status evaluation matrix, specifically: the average score is calculated according to the abnormal status evaluation matrix; and the operating status of each battery is determined according to the size relationship between all elements in the abnormal status evaluation matrix and the average score.

[0198] Optionally, the operating status of each battery is judged based on the size relationship between all elements in the abnormal status evaluation matrix and the mean score, specifically: if the abnormal status evaluation matrix value of the element at the (i, j)th position is not greater than the mean score, then the battery operating status result of the i-th battery in the j-th series battery branch is judged to be normal; if the abnormal status evaluation matrix value of the element at the (i, j)th position is greater than the mean score, then the battery operating status result of the i-th battery in the j-th series battery branch is judged to be abnormal.

[0199] In this embodiment, the abnormal state evaluation matrix S of all m×n batteries in step 104 is counted. c (t) is used to determine the abnormal state of each battery and locate the abnormal battery. The judgment process is explained using the mean shift as an example. Assume that the abnormal state evaluation matrix S c The element at position (i, j) in (t) is S i,j , then the abnormal state evaluation matrix S c The mean score of (t) is S mean , calculate the abnormal state evaluation matrix S c The average of all elements and scores in (t) is S mean The size relationship; if the element at position (i, j) is S i,j ≤S mean , then it is determined that the battery at the (i, j)th position (the i-th battery in the j-th series battery branch) has no abnormality; if the element at the (i, j)th position is S i,j >S mean , then it is determined that the battery at the (i, j)th position (the i-th battery in the j-th series battery branch) may be abnormal and requires special attention. The larger the deviation and the longer the duration, the more significant the abnormal state. Appropriate protective measures can be taken to complete the abnormal detection and positioning of the battery operating status.

[0200] In implementing the embodiment of the present invention, the dissimilarity matrix of each battery in the same series connection mode is calculated based on the real-time voltage value and temperature value of the series-connected batteries, and then the abnormality probability vector of each battery is obtained. The abnormality probability vector obtained based on the voltage value and temperature value is integrated by means of evidence fusion. At the same time, based on the current value of each battery string in parallel mode, the dissimilarity matrix of each battery string is calculated to obtain the abnormality probability vector of each battery string. The abnormality probability of each battery string is multiplied by the internal fusion abnormality probability vector of each battery string, and finally the abnormality probability score of each battery in the series-parallel structure of the battery energy storage system is obtained, realizing the abnormality detection result and positioning method of each battery operating state. Compared with the existing battery energy storage operating state abnormality detection and positioning method, the present invention integrates the real-time voltage, current and temperature original information of all batteries (features are not extracted here) to ensure the real-time detection. In addition, the present invention integrates the voltage and temperature information of each battery in the same battery string, and at the same time integrates the current information of each series battery branch, which can effectively improve the accuracy of battery operation status abnormality detection and form a precise score for each battery operation abnormality. This is conducive to the identification and location of each battery operation abnormality in the battery energy storage system, and then forms an effective early warning. This will enhance the safety level of the battery energy storage system operation and promote the more extensive, reliable and safe application of battery energy storage systems in new power systems.

[0201] Example 2

[0202] Accordingly, see Figure 5 , Figure 5 This is a structural diagram of a second embodiment of a detection system for the operating status of an energy storage battery provided by the present invention. Figure 5 As shown, the detection system for the operating status of the energy storage battery includes a series abnormality module 501, an evidence fusion module 502, a parallel abnormality module 503, a weighting module 504 and a status evaluation module 505;

[0203] The series abnormality module 501 is used to obtain the real-time data of each battery in the battery energy storage system and the series-parallel relationship of the batteries, and calculate the abnormality probability vector under each series structure based on the series structure and the real-time data of each battery. The real-time data includes real-time voltage data, real-time temperature data, and real-time current data, the abnormality probability vector includes a voltage abnormality probability vector and a temperature abnormality probability vector, and the series-parallel relationship of the batteries includes a series structure and a parallel structure.

[0204] The evidence fusion module 502 is used to fuse the voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure to obtain a series anomaly probability matrix;

[0205] The parallel abnormality module 503 is used to calculate the current abnormality probability vector of each battery string in the parallel structure based on the real-time current data of each battery string in the parallel structure;

[0206] The weighting module 504 is used to assign the current abnormality probability vector of each battery string in the parallel structure to the series abnormality probability matrix to obtain the abnormal state evaluation matrix;

[0207] The state evaluation module 505 is used to determine the operating state of each battery according to the abnormal state evaluation matrix and obtain the operating state detection result of each battery in the battery energy storage system; wherein the operating state detection result includes no abnormality and abnormal location.

[0208] By implementing an embodiment of the present invention, abnormal operating conditions of energy storage batteries can be detected and located. First, the series and parallel relationships of all batteries in the battery energy storage system are determined and numbered. Second, the inter-battery voltage deviation matrix is ​​calculated based on the real-time voltage data of each battery in the same battery string, thereby obtaining a voltage anomaly probability vector. Next, the inter-battery temperature deviation matrix and temperature anomaly probability vector are calculated based on the temperature data. Then, the voltage and temperature anomaly probability vectors of the batteries in the same battery string are integrated using evidence theory to form a series anomaly probability vector. Subsequently, the current test data of each parallel battery string is used to calculate the parallel branch current deviation matrix to obtain the current anomaly probability vector of each battery string. Next, the series anomaly probability vector is weighted using the current anomaly probability vector to obtain an anomaly score matrix for all batteries. Finally, the anomaly score matrix is ​​statistically analyzed to determine the abnormal state and location of each battery, thereby completing anomaly detection and location. Starting from multiple perspectives such as the voltage and temperature data of series batteries and the current data of parallel battery strings, a method for detecting and locating abnormal battery operation status in a hybrid battery energy storage system is established. Its unique feature is that it integrates the early warning information of the voltage and temperature differences of the series batteries, and then combines it with the basic information of the current difference of each series branch to form a comprehensive method for detecting and locating abnormal battery energy storage operation status. This improves the accuracy of the early warning of abnormal battery energy storage status and reduces the safety hazards in its application process. At the same time, it promotes the development of intelligent operation and maintenance technology for battery energy storage systems, expands the application scope of battery energy storage systems in new power systems, and contributes to my country's energy structure reform process.

[0209] The above specific embodiments further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for detecting the operating status of an energy storage battery, characterized in that: include: Acquire real-time data of each battery in the battery energy storage system and the battery series-parallel relationship, and calculate an abnormality probability vector under each series structure based on the series structure and the real-time data of each battery; wherein the real-time data includes real-time voltage data, real-time temperature data, and real-time current data, the abnormality probability vector includes a voltage abnormality probability vector and a temperature abnormality probability vector, and the battery series-parallel relationship includes the series structure and the parallel structure; Performing evidence fusion on the voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure to obtain a series anomaly probability matrix; Calculating a current abnormality probability vector of each battery string in the parallel structure according to the real-time current data of each battery string in the parallel structure; Assigning the current abnormality probability vector of each battery string in the parallel structure to the series abnormality probability matrix to obtain an abnormal state evaluation matrix; Determine the operating status of each battery according to the abnormal status evaluation matrix, and obtain an operating status detection result of each battery in the battery energy storage system; wherein the operating status detection result includes no abnormality and abnormal location; The abnormal probability vector under each series structure is calculated based on the real-time data of the series structure and each battery, specifically: the deviation matrix under the current series structure is calculated based on the absolute value of the difference between the real-time data of each battery under the current series structure and the real-time data of two adjacent batteries; wherein the deviation matrix includes a voltage deviation matrix and a current deviation matrix; based on the deviation matrix under the current series structure, the deviation matrices under several series structures are obtained; and based on the deviation matrix under each series structure, the abnormal probability vector under each series structure is calculated.

2. The method for detecting the operating status of an energy storage battery according to claim 1, wherein: The real-time data of each battery in the battery energy storage system and the battery series-parallel relationship are obtained as follows: Measuring the real-time data of each battery in the battery energy storage system by using sensors to obtain the real-time data of each battery; wherein the sensors include voltage sensors, temperature sensors and current sensors; According to the integrated structure of the battery energy storage system, the battery string and parallel connection situation is determined, and according to the battery string and parallel connection situation, the battery strings are numbered in a two-dimensional array to determine the number of each battery; wherein the battery string and parallel connection situation includes the number of each battery in the battery string, the number of parallel battery strings, and the battery string and parallel connection relationship.

3. The method for detecting the operating status of an energy storage battery according to claim 2, wherein: The calculation of the abnormal probability vector under each of the series structures according to the deviation matrix under each of the series structures is specifically as follows: Calculating the affinity matrix under the current series structure according to the deviation matrix under the current series structure and a preset similarity entropy value range; Normalizing the affinity matrix under the current series structure to obtain a probability matrix under the current series structure; wherein the probability matrix includes a voltage probability matrix and a temperature probability matrix; According to the probability matrix under the current series structure, the abnormal state scoring vector of each battery in the current series structure is calculated, wherein the abnormal state scoring vector includes a voltage abnormal state scoring vector and a temperature abnormal state scoring vector, and the formula is as follows: Among them, c U i,s(t) is the voltage abnormality state score vector of the sth battery in the ith battery string at time t, b U i,j,s(t) is the voltage probability matrix value of the sth battery under the i-th battery string at time t; Among them, c Tem i,s(t) is the temperature abnormality state score vector of the sth battery under the ith battery string at time t, b Tem i,j,s(t) is the temperature probability matrix value of the sth battery under the i-th battery string at time t; Normalizing the abnormal state score vector of each battery in the current series structure to obtain an abnormal probability vector in the current series structure; According to the abnormal probability vector under the current series structure, the abnormal probability vector under each series structure is obtained, and the formula is as follows: P U i(t)=[p U i,1 (t),p U i,2 (t),…,p U i,m (t)] T Among them, p U i,s(t) is the voltage abnormality probability vector of the sth battery under the ith battery string at time t, P U i(t) is the voltage abnormality probability vector of the i-th battery string at time t; P Tem i(t)=[p Tem i,1(t),p Tem i,2(t),…,p Tem i,m(t)] T Among them, p Tem i,s(t) is the temperature anomaly probability vector of the sth battery under the ith battery string at time t, P Tem i(t) is the temperature anomaly probability vector of the i-th battery string at time t.

4. The method for detecting the operating status of an energy storage battery according to claim 3, wherein: The affinity matrix under the current series structure is calculated according to the deviation matrix under the current series structure and the preset similarity entropy value range, specifically: Obtaining a deviation value of each battery according to the deviation matrix under the current series structure; wherein the deviation value includes a voltage deviation value and a temperature deviation value; The standard deviation of each battery is calculated based on the deviation value of each battery, the preset similarity entropy value range, the similarity entropy relationship, and the similarity deviation relationship through a binary search method, wherein the similarity entropy relationship includes a voltage similarity entropy relationship and a temperature similarity entropy relationship. The voltage similarity entropy relationship is expressed as follows: Among them, H(d U i,s(t)) is the voltage similarity entropy between the sth battery and the rth battery in the i-th battery string at time t, d U i,r|i,s (t) is the voltage similarity between the sth battery and the rth battery in the i-th battery string at time t; The temperature similarity entropy relationship is as follows: Among them, H(d Tem i,s(t)) is the temperature similarity entropy between the sth battery and the rth battery in the i-th battery string at time t, d Tem i,r|i,s (t) is the temperature similarity between the sth battery and the rth battery in the i-th battery string at time t; The similarity deviation relationship includes a voltage similarity deviation relationship and a temperature similarity deviation relationship. The voltage similarity deviation relationship is expressed as follows: Among them, σ U i,s (t) is the voltage standard deviation of the sth battery in the ith battery string at time t, ΔU i,s,r (t) is the voltage deviation between the sth battery and the rth battery in the i-th battery string at time t, ΔU i,s,k (t) is the voltage deviation value of the sth battery and the kth battery in the i-th battery string at time t; The temperature similarity deviation relationship is as follows: Among them, σ Tem i,s (t) is the temperature standard deviation of the sth battery in the ith battery string at time t, ΔT i,s,r (t) is the temperature deviation value of the sth battery and the rth battery in the i-th battery string at time t, ΔT i,s,k (t) is the temperature deviation value of the sth battery and the kth battery in the i-th battery string at time t; According to the similarity deviation relationship, the standard deviation of each battery in the current series structure, and the deviation value, the affinity matrix in the current series structure is calculated. The affinity matrix includes a voltage affinity matrix and a temperature affinity matrix. The formula is as follows: Among them, A U i(t) is the voltage affinity matrix of the i-th battery string at time t, is the voltage similarity between the jth battery and the mth battery in the i-th battery string at time t; Among them, A Tem i(t) is the temperature affinity matrix of the i-th battery string at time t, is the temperature similarity between the j-th battery and the m-th battery in the i-th battery string at time t.

5. The method for detecting the operating status of an energy storage battery according to claim 1, wherein: The voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure are subjected to evidence fusion to obtain a series anomaly probability matrix, which is specifically: The voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure are subjected to evidence fusion to obtain a series anomaly probability matrix under the same series structure. The evidence fusion formula is as follows: P S i(t)=[P S i,1(t),P S i,2(t),…,P S i,m(t)] T Among them, P S i(t) is the probability vector of abnormal connection of the i-th battery string at time t, The voltage anomaly probability vector and the temperature anomaly probability vector of the i-th battery string at time t are fused together to form the B-th battery string. i,r The probability of battery abnormality, is the Bth battery string under the i-th battery string at time t i,r The voltage abnormality probability vector of the battery, is the Bth battery string under the i-th battery string at time t i,r The temperature abnormality probability vector of the battery, k c The conflict coefficient is equal to According to the series anomaly probability matrix under the same series structure, the series anomaly probability matrix is ​​obtained, and the formula is as follows: P S (t)=[P S 1(t),P S 1(t),…,P S n (t)]∈R m×n Among them, P S (t) is the series abnormality probability vector.

6. The method for detecting the operating status of an energy storage battery according to claim 5, wherein: The current abnormality probability vector of each battery string in the parallel structure is weighted to the series abnormality probability matrix to obtain an abnormal state evaluation matrix, which is specifically: The current abnormality probability vectors of each battery string in the parallel structure are weighted by column to the series abnormality probability matrix to obtain an abnormal state evaluation matrix, specifically: S C (t)=[p I 1(t)×P1 S (t),p I 2(t)×P2 S (t),…,p I n (t)×P n S (t)]∈R m×n Among them, S C (t) is the abnormal state evaluation matrix, P S n(t) is the series abnormality probability vector of the nth battery string at time t, p I n (t) The current abnormality probability vector of the nth battery string at time t.

7. The method for detecting the operating status of an energy storage battery according to claim 1, wherein: The operation status of each battery is determined according to the abnormal state evaluation matrix, specifically: Calculating a score mean according to the abnormal state evaluation matrix; The operating state of each battery is determined according to the size relationship between all elements in the abnormal state evaluation matrix and the score mean.

8. The method for detecting the operating status of an energy storage battery according to claim 7, wherein: The operation state of each battery is judged according to the size relationship between all elements in the abnormal state evaluation matrix and the mean score, specifically: If the abnormal state evaluation matrix value of the element at the (i, j)th position is not greater than the score mean, then the battery operation state result of the i-th battery in the j-th series battery branch is determined to be normal; If the abnormal state evaluation matrix value of the element at the (i, j)th position is greater than the score mean, it is determined that the battery operating state of the i-th battery in the j-th series battery branch is abnormal.

9. A detection system for the operating status of an energy storage battery, characterized in that: include: Series anomaly module, evidence fusion module, parallel anomaly module, empowerment module and status evaluation module; The series abnormality module is used to obtain the real-time data of each battery in the battery energy storage system and the battery series-parallel relationship, and calculate the abnormal probability vector under each series structure according to the series structure and the real-time data of each battery; wherein the real-time data includes real-time voltage data, real-time temperature data and real-time current data, the abnormal probability vector includes a voltage abnormality probability vector and a temperature abnormality probability vector, and the battery series-parallel relationship includes the series structure and the parallel structure; The evidence fusion module is used to fuse the voltage anomaly probability vector and the temperature anomaly probability vector under the same series structure to obtain a series anomaly probability matrix; The parallel abnormality module is used to calculate the current abnormality probability vector of each battery string in the parallel structure according to the real-time current data of each battery string in the parallel structure; The weighting module is used to assign the current abnormality probability vector of each battery string in the parallel structure to the series abnormality probability matrix to obtain an abnormal state evaluation matrix; The state evaluation module is used to determine the operating state of each battery according to the abnormal state evaluation matrix, and obtain the operating state detection result of each battery in the battery energy storage system; wherein the operating state detection result includes no abnormality and abnormal location; The abnormal probability vector under each series structure is calculated based on the real-time data of the series structure and each battery, specifically: the deviation matrix under the current series structure is calculated based on the absolute value of the difference between the real-time data of each battery under the current series structure and the real-time data of two adjacent batteries; wherein the deviation matrix includes a voltage deviation matrix and a current deviation matrix; based on the deviation matrix under the current series structure, the deviation matrices under several series structures are obtained; and based on the deviation matrix under each series structure, the abnormal probability vector under each series structure is calculated.

Images