Power battery voltage fault online diagnosis processing method based on entropy algorithm

Abstract

A power battery voltage fault online diagnosis processing method based on an entropy algorithm, comprising: acquiring a voltage time series value of each cell of a power battery to be diagnosed of an electric vehicle to establish a voltage data matrix (S100); on the basis of a sliding window preset with a constant length, sliding downwards across a voltage matrix, and on the basis of a data processing principle, evicting and reconstructing data within the sliding window each time (S102); then calculating a voltage entropy value of each cell, performing Z-score conversion on the calculation results to obtain anomaly assessment coefficients, and performing outlier identification on the coefficients (S105); if there is an outlier, locating the number and time of an abnormal cell, assessing the degree of abnormality, and transmitting a corresponding prompt to a driver on the basis of a safety management strategy at the same time (S107); and if there is no outlier, a time window continuing to slide downwards, and repeating the described processes to perform iterative calculation (S108). The method can accurately locate the occurrence time and location of cell voltage faults, thereby improving the safety of electric vehicles to a certain extent.

Description

An online diagnosis and processing method for power battery voltage fault based on entropy algorithm

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to the Chinese patent application with application number 202311016786.1 filed with the Patent Office of China on August 14, 2023, entitled “A method for online diagnosis and processing of power battery voltage fault based on entropy algorithm”, the entire contents of which are incorporated by reference into this application. Technical Field

[0003] The present invention relates to the technical field of battery fault diagnosis, and in particular to an online diagnosis and processing method for power battery voltage fault based on an entropy algorithm. Background Art

[0004] In recent years, the electric vehicle industry has experienced rapid growth, and vehicle accidents have become commonplace. Power batteries are often blamed for this, making fault diagnosis of electric vehicle power batteries crucial. Power battery system failures can often be traced back to a specific battery cell or cells. Cell failures primarily include overvoltage, undervoltage, short circuit, overtemperature, and poor consistency. These failures are readily apparent in the cell voltage parameter, primarily manifesting as inconsistent voltage fluctuations. Therefore, rapid and accurate monitoring and assessment of abnormal cell voltage fluctuations can provide a more comprehensive understanding of power battery system safety. Various entropy methods are widely used in battery fault diagnosis due to their ability to effectively assess system disorder. However, these methods sacrifice computational efficiency for model accuracy.

[0005] Based on this, the present invention proposes an online diagnosis and processing method for power battery voltage fault based on entropy algorithm.

[0006] Summary of the Invention

[0007] In order to make up for the deficiencies of the prior art and solve the above-mentioned technical problems, the present invention provides an online diagnosis and processing method for power battery voltage fault based on an entropy algorithm.

[0008] A method for online diagnosis and processing of power battery voltage faults based on an entropy algorithm, the method comprising the following steps:

[0009] Step 100: Obtain the time series value of the voltage data of each single cell in the battery pack of the vehicle to be diagnosed, and construct a voltage data matrix A, where:

[0010] Where m represents the time sequence length, n represents the total number of single cells in the battery pack, and a i,jIndicates the voltage value of the j-th single cell at the i-th index, i = 1, 2, ..., m; j = 1, 2, ..., n;

[0011] Step 101: Preset the length and width of the initial sliding time window, and use the sliding window to intercept data from the voltage data matrix region by region to form an intercepted voltage matrix B;

[0012] Step 102: extract and reconstruct the voltage values ​​in the window according to data processing principles to form a new voltage matrix D;

[0013] Step 103: Calculate the Shannon entropy value of each single battery using an optimized entropy algorithm based on the reconstructed data;

[0014] Step 104: converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain a monomer voltage abnormal fluctuation assessment coefficient;

[0015] Step 105: Identify abnormal monomers based on the security assessment strategy;

[0016] Step 106: Determine whether there is an abnormal monomer. If yes, go to step 107; otherwise, go to step 108.

[0017] Step 107: If there is an abnormal value, locate the position and occurrence time of the abnormal unit, further determine the degree of abnormality, and send a corresponding prompt to the driver;

[0018] Step 108: If there is no outlier, the time window moves downward and steps 101 to 106 are repeated for the next iterative calculation.

[0019] Preferably, in step 101, the length of the initial sliding time window is preset to be k, and the width is the same as the total number of single cells, n. The voltage data matrix intercepted by each sliding window is B, where:

[0020] in

[0021] Preferably, the method for processing the voltage value data in the sliding window in step 102 is specifically as follows:

[0022] S200: Real-time monitoring of the voltage difference of each single cell at consecutive moments, i.e., calculating the voltage difference matrix C of the voltage data of each single cell at consecutive moments in the time window matrix B, where:

[0023] where c i,j represents the difference between the voltage value of the j-th single cell at the i+1th index and the voltage value at the i-th index, i=1, 2, ..., k-1; j=1, 2, ..., n;

[0024] S201: Find the row indexes of all data in the matrix C where the voltage difference of each single battery is less than or equal to 0.001V and exceeds k / 10 consecutively;

[0025] S202: Kick out the data segments corresponding to the row indices in matrix B that meet the conditions in S201;

[0026] S203: Splice the remaining data in time sequence, take the mean of the two end point data at the splicing point, and form a new voltage matrix D, where:

[0027] in

[0028] Preferably, in step 103, the Shannon entropy value of each single battery is calculated using the optimized entropy algorithm based on the reconstructed data matrix D. The specific steps are:

[0029] S300: Find the maximum and minimum values ​​in the time window, that is, find the minimum value d in the matrix D min and the maximum value d max for:

[0030] S301: The area formed by the two maximum values ​​(d min , d max ) is divided into l intervals:

[0031] Where L = 1, 2,…, l.

[0032] S302: Calculate the frequency matrix F of each single cell voltage data in the matrix D falling into the above-mentioned l intervals, and obtain the frequency matrix F, where:

[0033] where f i,j Indicates that all data of single battery j in matrix D fall into the interval The number of .

[0034] S303: Divide the total number by the frequency to obtain the regional probability matrix P of each individual data in each region, where

[0035] in The probability that the voltage data of the single cell j in the matrix D falls into the i-th interval, and the Shannon entropy of each single cell calculated by the matrix B is: H(B)=[H1,H2,L,H j ,L,H n ]

[0036] S304: According to Calculate the Shannon entropy value of each cell, that is, the Shannon entropy value of the j-th cell calculated from the time window B.

[0037] Preferably, in step 104, the specific steps of converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain the monomer voltage abnormal fluctuation assessment coefficient are as follows:

[0038] S400: Obtain the Shannon entropy sequence H(B) of each single cell.

[0039] S401: Calculate the mean value μ of the Shannon entropy of each single cell H , the calculation formula is:

[0040] S402: Kick out H(B) with more than 2*μ H After the data is collected, the standard deviation of the Shannon entropy sequence of the remaining cells is calculated; that is, the cells with more than 2*μ in the Shannon entropy sequence H(B) are removed. H Shannon entropy value, forming a new Shannon entropy sequence, Calculate the mean and standard deviation of the sequence H(B)' using the following formulas:

[0041] S403: Calculate the median value H of the Shannon entropy of each single cell me ; That is, calculate the median H of the Shannon entropy sequence H(B) of each single cell me ;

[0042] S404: According to Normalize the Shannon entropy value of each single cell; that is, calculate the abnormality assessment coefficient of each single cell in the current time window based on the Z score theory.

[0043] Preferably, the safety assessment strategy for outliers in step 105 is:

[0044] S500: Identify a single cell with a voltage of 3.5 < |AF| in each iterative calculation of the time window as a faulty cell with abnormal voltage fluctuation;

[0045] S501: If |AF|≤3.5, execute S502 and S503;

[0046] S502: No voltage fault alarm;

[0047] S503: Perform next iterative calculation;

[0048] S504: If 3.5<|AF|≤4, execute S505 and S506, or determine whether to execute S510 and S510, or S503 according to S507;

[0049] S505: Voltage level 2 fault alarm;

[0050] S506: Calculate the duration of the secondary fault;

[0051] S507: Whether the threshold is exceeded;

[0052] S508: If 4<|AF|, execute S509 and S510;

[0053] S509: voltage level 1 fault alarm;

[0054] S510: The vehicle needs to be inspected immediately.

[0055] Preferably, in step 108, if there is no outlier, the time window moves downward once, and steps 101 to 106 are repeated to perform the next iterative calculation.

[0056] An online diagnostic and processing device for power battery voltage faults based on an entropy algorithm includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the following steps are implemented:

[0057] Obtain the time series value of the voltage data of each single cell in the battery pack of the vehicle to be diagnosed and construct the voltage data matrix A;

[0058] Step 101: Preset the length and width of the initial sliding time window, and use the sliding window to intercept data from the voltage data matrix region by region to form an intercepted voltage matrix B;

[0059] Step 102: extract and reconstruct the voltage values ​​in the window according to data processing principles to form a new voltage matrix D;

[0060] Step 103: Calculate the Shannon entropy value of each single battery using an optimized entropy algorithm based on the reconstructed data;

[0061] Step 104: converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain a monomer voltage abnormal fluctuation assessment coefficient;

[0062] Step 105: Identify abnormal monomers based on the security assessment strategy;

[0063] Step 106: Determine whether there is an abnormal monomer. If yes, go to step 107; otherwise, go to step 108.

[0064] Step 107: If there is an abnormal value, locate the position and occurrence time of the abnormal unit, further determine the degree of abnormality, and send a corresponding prompt to the driver;

[0065] Step 108: If there is no outlier, the time window moves downward and steps 101 to 106 are repeated for the next iterative calculation.

[0066] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:

[0067] Step 100: Obtain the time series value of the voltage data of each single cell in the battery pack of the vehicle to be diagnosed, and construct a voltage data matrix A;

[0068] Step 101: Preset the length and width of the initial sliding time window, and use the sliding window to intercept data from the voltage data matrix region by region to form an intercepted voltage matrix B;

[0069] Step 102: extract and reconstruct the voltage values ​​in the window according to data processing principles to form a new voltage matrix D;

[0070] Step 103: Calculate the Shannon entropy value of each single battery using an optimized entropy algorithm based on the reconstructed data;

[0071] Step 104: converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain a monomer voltage abnormal fluctuation assessment coefficient;

[0072] Step 105: Identify abnormal monomers based on the security assessment strategy;

[0073] Step 106: Determine whether there is an abnormal monomer. If yes, go to step 107; otherwise, go to step 108.

[0074] Step 107: If there is an abnormal value, locate the position and occurrence time of the abnormal unit, further determine the degree of abnormality, and send a corresponding prompt to the driver;

[0075] Step 108: If there is no outlier, the time window moves downward and steps 101 to 106 are repeated for the next iterative calculation.

[0076] The beneficial effects of the present invention are as follows:

[0077] 1. Compared with the existing battery fault diagnosis method based on Shannon entropy, the beneficial effects of the present invention are:

[0078] Since this method uses an entropy calculation method based on interval probability, and the interval changes with the voltage data, the size of the Shannon entropy is easily affected by multiple voltage data segments with a long duration and no fluctuations, resulting in a high misdiagnosis rate. Based on this, the present invention adds the steps of kicking out and reconstructing the original data. The data processing principle ensures that the kicked out data does not contain fault data with abnormal fluctuations, and at the same time kicks out data segments that are prone to misdiagnosis, greatly improving the accuracy of fault diagnosis.

[0079] The mean and standard deviation in the original Z-score method are easily affected by the extreme value data during calculation. In order to further improve the accuracy of the method in identifying abnormal monomers, the mean is replaced by the median during the Shannon entropy standardization process in this paper, and the outliers are first removed when calculating the standard deviation, which further reduces the misdiagnosis rate.

[0080] The present invention uses the sliding time window principle to locate the position of the abnormal monomer and the time when the abnormality occurred by calculating the voltage data in the sliding window, reducing the amount of calculation and realizing online application. The method is different from the existing fault diagnosis method that calculates data of the entire life cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] The present invention will be further described below with reference to the accompanying drawings.

[0082] FIG1 is a schematic flow chart of an online diagnosis and processing method for power battery voltage fault based on an entropy algorithm provided by the present invention;

[0083] FIG2 is a schematic diagram of a data ejection and reconstruction principle flow chart provided by the present invention;

[0084] FIG3 is a schematic diagram of the process flow of the optimized entropy algorithm provided by the present invention;

[0085] FIG4 is a schematic diagram of the optimized Z-score algorithm provided by the present invention;

[0086] FIG5 is a schematic diagram of the process flow of abnormal value diagnosis processing provided by the present invention. DETAILED DESCRIPTION

[0087] The technical solution of the present application is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. Unless there is a conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0088] A method for online diagnosis and processing of power battery voltage faults based on an entropy algorithm, the method comprising the following steps:

[0089] Step 100: Obtain the time series value of the voltage data of each single cell in the battery pack of the vehicle to be diagnosed, and construct a voltage data matrix A, where:

[0090] Among them, m represents the time sequence length, n represents the total number of single cells in the battery pack, and a i,j Indicates the voltage value of the j-th single cell at the i-th index, i = 1, 2, ..., m; j = 1, 2, ..., n;

[0091] Step 101: Preset the length and width of the initial sliding time window, and use the sliding window to intercept data from the voltage data matrix region by region to form an intercepted voltage matrix B;

[0092] Step 102: extract and reconstruct the voltage values ​​in the window according to data processing principles to form a new voltage matrix D;

[0093] Step 103: Calculate the Shannon entropy value of each single battery using an optimized entropy algorithm based on the reconstructed data;

[0094] Step 104: converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain a monomer voltage abnormal fluctuation assessment coefficient;

[0095] Step 105: Identify abnormal monomers based on the security assessment strategy;

[0096] Step 106: Determine whether there is an abnormal monomer. If yes, go to step 107; otherwise, go to step 108.

[0097] Step 107: If there is an abnormal value, locate the position and occurrence time of the abnormal unit, further determine the degree of abnormality, and send a corresponding prompt to the driver;

[0098] Step 108: If there is no outlier, the time window moves downward and steps 101 to 106 are repeated for the next iterative calculation.

[0099] In step 101, the length of the initial sliding time window is preset to be k, and the width is the same as the total number of single cells, n. The voltage data matrix intercepted by each sliding window is B, where:

[0100] in

[0101] The method for processing the voltage value data in the sliding window in step 102 is specifically as follows:

[0102] S200: Real-time monitoring of the voltage difference of each single cell at consecutive moments, i.e., calculating the voltage difference matrix C of the voltage data of each single cell at consecutive moments in the time window matrix B, where:

[0103] where c i,j represents the difference between the voltage value of the j-th single cell at the i+1th index and the voltage value at the i-th index, i=1, 2, ..., k-1; j=1, 2, ..., n;

[0104] S201: Find the row indexes of all data in the matrix C where the voltage difference of each single battery is less than or equal to 0.001V and exceeds k / 10 consecutively;

[0105] S202: Kick out the data segments corresponding to the row indices in matrix B that meet the conditions in S201;

[0106] S203: Splice the remaining data in time sequence, take the mean of the two end point data at the splicing point, and form a new voltage matrix D, where:

[0107] in

[0108] In step 103, the Shannon entropy value of each single battery is calculated using the optimized entropy algorithm based on the reconstructed data matrix D. The specific steps are:

[0109] S300: Find the maximum and minimum values ​​in the time window, that is, find the minimum value d in the matrix D min and the maximum value d max for:

[0110] S301: The area formed by the two maximum values ​​(d min , d max ) is divided into l intervals:

[0111] Where L = 1, 2,…, l.

[0112] S302: Calculate the frequency matrix F of each single cell voltage data in the matrix D falling into the above-mentioned l intervals, and obtain the frequency matrix F, where:

[0113] where f i,j Indicates that all data of single battery j in matrix D fall into the interval The number of .

[0114] S303: Divide the total number by the frequency to obtain the regional probability matrix P of each individual data in each region, where

[0115] in The probability that the voltage data of the single cell j in the matrix D falls into the i-th interval, and the Shannon entropy of each single cell calculated by the matrix B is: H(B)=[H1,H2,L,H j ,L,H n ]

[0116] S304: According to Calculate the Shannon entropy value of each cell, that is, the Shannon entropy value of the j-th cell calculated from the time window B.

[0117] In step 104, the Shannon entropy value of each monomer is converted based on the optimized Z-score method to obtain the monomer voltage abnormal fluctuation assessment coefficient. The specific steps are as follows:

[0118] S400: Obtain the Shannon entropy sequence H(B) of each single cell.

[0119] S401: Calculate the mean value μ of the Shannon entropy of each single cell H , the calculation formula is:

[0120] S402: Kick out H(B) with more than 2*μ H After the data is collected, the standard deviation of the Shannon entropy sequence of the remaining cells is calculated; that is, the cells with more than 2*μ in the Shannon entropy sequence H(B) are removed. H Shannon entropy value, forming a new Shannon entropy sequence H(B)'=[H1,H2,L,H g ], Calculate the mean and standard deviation of the sequence H(B)' using the following formulas:

[0121] S403: Calculate the median value H of the Shannon entropy of each single cell me ; That is, calculate the median H of the Shannon entropy sequence H(B) of each single cell me ;

[0122] S404: According to Normalize the Shannon entropy value of each single cell; that is, calculate the abnormality assessment coefficient of each single cell in the current time window based on the Z score theory.

[0123] The safety assessment strategy for outliers in step 105 is:

[0124] S500: Identify a single cell with a voltage of 3.5 < |AF| in each iterative calculation of the time window as a faulty cell with abnormal voltage fluctuation;

[0125] S501: If |AF|≤3.5, execute S502 and S503;

[0126] S502: No voltage fault alarm;

[0127] S503: Perform next iterative calculation;

[0128] S504: If 3.5<|AF|≤4, execute S505 and S506, or determine whether to execute S510 and S510, or S503 according to S507;

[0129] S505: Voltage level 2 fault alarm;

[0130] S506: Calculate the duration of the secondary fault;

[0131] S507: Whether the threshold is exceeded;

[0132] S508: If 4<|AF|, execute S509 and S510;

[0133] S509: voltage level 1 fault alarm;

[0134] S510: The vehicle needs to be inspected immediately;

[0135] In step 108 , if there is no outlier, the time window moves downward once, and steps 101 to 106 are repeated to perform the next iterative calculation.

[0136] Specifically, assuming that there are 95 single cells in the battery box of an electric vehicle to be tested, 200 sampling data are selected to establish the time series matrix A of all single cells as follows:

[0137] where a i,j Indicates the voltage value of the j-th single cell at the i-th index, i = 1, 2, ..., 200; j = 1, 2, ..., 95;

[0138] In step S101, the length of the initial sliding window is preset to be 100 and the width is preset to be 95. The sliding window is used to intercept data from the voltage time series matrix region by region. The voltage data matrix B intercepted for the first time is:

[0139] Based on the voltage data matrix B, step S102 is performed according to the process shown in FIG2 , that is, the data processing principle is specifically as follows:

[0140] In step S200, the voltage difference of each single cell at consecutive moments is monitored in real time, that is, the voltage difference matrix C of the voltage data of each single cell at consecutive moments in the time sliding window matrix B is calculated:

[0141] In step S201, the row indexes of all data in the matrix C where the voltage difference of each single battery cell is less than or equal to 0.001V for more than 10 consecutive times are found;

[0142] In step S202, the data segments corresponding to the row indices in matrix B that meet the conditions of step S201 are kicked out;

[0143] In step S203, the remaining data are spliced ​​in sequence according to the time sequence, and the mean of the two end point data is taken at the splicing point to form a new voltage matrix D:

[0144] in

[0145] Step S103 is performed according to the process shown in FIG3 , that is, the optimized entropy value algorithm is specifically as follows:

[0146] In step S300, the minimum value d in the matrix D is obtained. min and the maximum value d max for:

[0147] In step S301, the area (d min , d max ) is divided into 10 intervals:

[0148] Where L = 1, 2,…, 10.

[0149] In step S302, the frequency of each single cell voltage data in the matrix D falling into the above 10 intervals is calculated to obtain the matrix F:

[0150] where f i,j Indicates that all data of single battery j in matrix D fall into the interval The number of.

[0151] In step S303, the probability matrix P is calculated based on the matrix F:

[0152] in It represents the probability that the voltage data of single cell j in matrix D falls into the i-th interval.

[0153] In step S304, the Shannon entropy of each battery cell calculated from the matrix B is: H(B) = [H1, H2, L, H j ,L,H 95 ]

[0154] in Represents the Shannon entropy value of the j-th single cell in time window B.

[0155] Step S104 is performed according to the process shown in FIG4 , that is, the optimized Z score algorithm is:

[0156] In step S401, the mean of the Shannon entropy sequence H(B) obtained in step S400 is calculated using the following formula:

[0157] In step S402, the Shannon entropy sequence H(B) with more than 2*μ is kicked out. H Shannon entropy value, forming a new Shannon entropy sequence H(B)'=[H1,H2,L,H g ], Calculate the mean and standard deviation of the sequence H(B)' using the following formulas:

[0158] In step S403, the median value H of the Shannon entropy sequence is calculated. me ;

[0159] In step S404, the abnormality assessment coefficient of each single battery in the current time window is calculated according to the Z score theory, and the calculation formula is:

[0160] Finally, steps S105, S106, S107 and S108 are executed according to the process shown in FIG5 , that is, when the power battery voltage fails, that is, an abnormal value occurs, the safety assessment strategy is as follows:

[0161] In step S501 , based on a large amount of real fault and normal vehicle data analysis and verification, a single cell with a voltage of 3.5<|AF| is identified as a faulty cell with abnormal voltage fluctuation.

[0162] In step S502, if |AF|≤3.5, there is no abnormal voltage fluctuation fault in the single cell, and the system does not have a voltage fault alarm. Step S503 is performed, that is, the next iterative calculation is performed. The voltage data matrix B intercepted for the second time is:

[0163] Then repeat steps S102-S106.

[0164] In step S504, if 3.5<|AF|≤4, the system issues a voltage level 2 fault alarm. In step S506, the level 2 fault alarm duration is calculated. If it exceeds the threshold, step S510 is executed; otherwise, step S503 is executed.

[0165] In step S508, if 4<|AF|, the system issues a voltage level 1 fault alarm, prompting the driver to stop the vehicle immediately and conduct an inspection.

[0166] The above process illustrates the complete calculation process when the sliding window intercepts data once, that is, the online diagnosis and processing method for power battery voltage fault based on the entropy algorithm provided by this embodiment of the present invention. When implemented online, the above steps are continuously repeated for iterative calculation to achieve real-time and accurate assessment of abnormal voltage fluctuation faults.

[0167] Based on the same inventive concept as the method for online diagnosis and processing of power battery voltage fault based on an entropy algorithm in the aforementioned embodiment, the present invention further provides a device for online diagnosis and processing of power battery voltage fault based on an entropy algorithm, on which a computer program is stored. When the program is executed by a processor, the steps of any one of the aforementioned methods for online diagnosis and processing of power battery voltage fault based on an entropy algorithm are implemented as follows:

[0168] It includes a bus architecture, which may include any number of interconnected buses and bridges that link various circuits including one or more processors represented by a processor and memory represented by a memory. The bus may also connect various other circuits such as peripherals, voltage regulators, sensors, cameras, and power management circuits, all of which are well known in the art and, therefore, will not be further described in this invention.

[0169] Example 3

[0170] Based on the same inventive concept as the entropy algorithm-based online diagnosis and processing method for power battery voltage faults in the aforementioned embodiment, the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the following steps are implemented:

[0171] Obtain the time series value of the voltage data of each single cell in the battery pack of the vehicle to be diagnosed and construct the voltage data matrix A;

[0172] Step 101: Preset the length and width of the initial sliding time window, and use the sliding window to intercept data from the voltage data matrix region by region to form an intercepted voltage matrix B;

[0173] Step 102: extract and reconstruct the voltage values ​​in the window according to data processing principles to form a new voltage matrix D;

[0174] Step 103: Calculate the Shannon entropy value of each single battery using an optimized entropy algorithm based on the reconstructed data;

[0175] Step 104: converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain a monomer voltage abnormal fluctuation assessment coefficient;

[0176] Step 105: Identify abnormal monomers based on the security assessment strategy;

[0177] Step 106: Determine whether there is an abnormal monomer. If yes, go to step 107; otherwise, go to step 108.

[0178] Step 107: If there is an abnormal value, locate the position and occurrence time of the abnormal unit, further determine the degree of abnormality, and send a corresponding prompt to the driver;

[0179] Step 108: If there is no outlier, the time window moves downward and steps 101 to 106 are repeated for the next iterative calculation.

[0180] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:

[0181] Step 100: Obtain the time series value of the voltage data of each single cell in the battery pack of the vehicle to be diagnosed, and construct a voltage data matrix A;

[0182] Step 101: Preset the length and width of the initial sliding time window, and use the sliding window to intercept data from the voltage data matrix region by region to form an intercepted voltage matrix B;

[0183] Step 102: extract and reconstruct the voltage values ​​in the window according to data processing principles to form a new voltage matrix D;

[0184] Step 103: Calculate the Shannon entropy value of each single battery using an optimized entropy algorithm based on the reconstructed data;

[0185] Step 104: converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain a monomer voltage abnormal fluctuation assessment coefficient;

[0186] Step 105: Identify abnormal monomers based on the security assessment strategy;

[0187] Step 106: Determine whether there is an abnormal monomer. If yes, go to step 107; otherwise, go to step 108.

[0188] Step 107: If there is an abnormal value, locate the position and occurrence time of the abnormal unit, further determine the degree of abnormality, and send a corresponding prompt to the driver;

[0189] Step 108: If there is no outlier, the time window moves downward and steps 101 to 106 are repeated for the next iterative calculation.

[0190] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0191] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0192] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0193] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0194] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. An online diagnosis and processing method for power battery voltage fault based on entropy algorithm, characterized by: The method comprises the following steps: Step 100: Obtain the voltage data time series value of each single cell in the battery pack of the vehicle to be diagnosed, and construct a voltage data matrix A, where: Where m represents the time sequence length, n represents the total number of single cells in the battery pack, and a i,j Indicates the voltage value of the j-th single cell at the i-th index, i = 1, 2, ..., m; j = 1, 2, ..., n; Step 101: Preset the length and width of the initial sliding time window, and use the sliding window to intercept data from the voltage data matrix region by region to form an intercepted voltage matrix B; Step 102: according to the data processing principle, the voltage values ​​in the window are kicked out and reconstructed to form a new voltage matrix D; Step 103: Calculate the Shannon entropy value of each single cell using an optimized entropy value algorithm based on the reconstructed data; Step 104: converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain a monomer voltage abnormal fluctuation assessment coefficient; Step 105: Identify abnormal monomers based on the security assessment strategy; Step 106: Determine whether there is an abnormal monomer, if yes, execute step 107, otherwise execute step 108; Step 107: If there is an abnormal value, locate the position and occurrence time of the abnormal monomer, further determine the degree of abnormality, and send a corresponding prompt to the driver; Step 108: If there is no outlier, the time window moves downward and steps 101 to 106 are repeated to perform the next iterative calculation.

2. The method for online diagnosis and processing of power battery voltage fault based on entropy algorithm according to claim 1 is characterized in that: In step 101, the length of the initial sliding time window is preset to be k, and the width is the same as the total number of single cells, n. The voltage data matrix intercepted by each sliding window is B, where: in 3. The method for online diagnosis and processing of power battery voltage fault based on entropy algorithm according to claim 2 is characterized in that: The method for processing the voltage value data in the sliding window in step 102 is specifically as follows: S200: Real-time monitoring of the voltage difference of each single cell at consecutive moments, that is, calculating the voltage difference matrix C of the voltage data of each single cell at consecutive moments in the time window matrix B, where: where c i,j represents the difference between the voltage value of the j-th single cell at the i+1th index and the voltage value of the i-th index, i=1,2,…,k-1; j=1,2,…,n; S201: Find the row indexes of all data in the matrix C where the voltage difference of each single cell is less than or equal to 0.001V and exceeds k / 10 consecutively; S202: Kick out the data fragments corresponding to the row indices in matrix B that meet the condition in S201; S203: splice the remaining data in time sequence, take the mean of the two end point data at the splicing point, and form a new voltage matrix D, where: in 4. The method for online diagnosis and processing of power battery voltage fault based on entropy algorithm according to claim 3 is characterized in that: In step 103, the Shannon entropy value of each single battery is calculated based on the reconstructed data matrix D using the optimized entropy value algorithm. The specific steps are: S300: Find the maximum and minimum values ​​in the time window, that is, find the minimum value d in the matrix D min and the maximum value d max for: S301: The area (d min , d max ) is divided into l intervals: Where L = 1, 2,…, l. S302: Calculate the frequency matrix F of each single cell voltage data in the matrix D falling into the above l intervals, and obtain the frequency matrix F, where: where f i,j Indicates that all data of single cell j in matrix D fall into the interval The number of S303: Divide the frequency by the total number to obtain the regional probability matrix P of each monomer data in each region, where in It represents the probability that the voltage data of the single cell j in the matrix D falls into the i-th interval, and the Shannon entropy of each single cell calculated by the matrix B is: H(B)=[H1,H2,L,H j ,L,H n ] S304: According to The Shannon entropy value of each monomer is calculated, which means the Shannon entropy value of the j-th monomer battery calculated by time window B.

5. The method for online diagnosis and processing of power battery voltage fault based on entropy algorithm according to claim 4 is characterized in that: In step 104, the Shannon entropy value of each monomer is converted based on the optimized Z-score method to obtain the monomer voltage abnormal fluctuation assessment coefficient. The specific steps are as follows: S400: Obtain the Shannon entropy sequence H(B) of each single cell. S401: Calculate the mean value μ of the Shannon entropy of each single cell H , the calculation formula is: S402: Kick out H(B) with more than 2*μ H After the data is collected, the standard deviation of the remaining single-cell Shannon entropy sequence is calculated; that is, the Shannon entropy sequence H(B) with more than 2*μ is kicked out H Shannon entropy values ​​form a new Shannon entropy sequence H(B)'=[H1,H2,L,H g ], Calculate the mean and standard deviation of the sequence H(B)' using the following formulas: S403: Calculate the median value H of the Shannon entropy of each single cell me ; That is, calculate the median H of the Shannon entropy sequence H(B) of each single cell me ; S404: According to Standardize the Shannon entropy value of each single cell; that is, calculate the abnormal evaluation coefficient of each single cell in the current time window according to the Z score theory.

6. The method for online diagnosis and processing of power battery voltage fault based on entropy algorithm according to claim 5 is characterized in that: The safety assessment strategy for outliers in step 105 is: S500: identifying a single cell with a voltage of 3.5 < |AF| in each iterative calculation of the time window as a faulty single cell with abnormal voltage fluctuation; S501: If |AF|≤3.5, execute S502 and S503; S502: No voltage fault alarm; S503: Perform next iterative calculation; S504: If 3.5<|AF|≤4, execute S505 and S506, or determine whether to execute S510 and S510, or S503 according to S507; S505: voltage level 2 fault alarm; S506: Calculate the duration of the secondary fault; S507: Whether the threshold is exceeded; S508: If 4<|AF|, execute S509 and S510; S509: voltage level 1 fault alarm; S510: The vehicle needs to be inspected immediately.

7. The method for online diagnosis and processing of power battery voltage fault based on entropy algorithm according to claim 6 is characterized in that: In step 108, if there is no abnormal value, the time window moves downward once, and steps 101 to 106 are repeated to perform the next iterative calculation.

8. An online diagnosis and processing device for power battery voltage fault based on entropy algorithm, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the program: Obtain the voltage data time series value of each single cell in the battery pack of the vehicle to be diagnosed, and construct a voltage data matrix A; Step 101: Preset the length and width of the initial sliding time window, and use the sliding window to intercept data from the voltage data matrix region by region to form an intercepted voltage matrix B; Step 102: according to the data processing principle, the voltage values ​​in the window are kicked out and reconstructed to form a new voltage matrix D; Step 103: Calculate the Shannon entropy value of each single cell using an optimized entropy value algorithm based on the reconstructed data; Step 104: converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain a monomer voltage abnormal fluctuation assessment coefficient; Step 105: Identify abnormal monomers based on the security assessment strategy; Step 106: Determine whether there is an abnormal monomer, if yes, execute step 107, otherwise execute step 108; Step 107: If there is an abnormal value, locate the position and occurrence time of the abnormal monomer, further determine the degree of abnormality, and send a corresponding prompt to the driver; Step 108: If there is no outlier, the time window moves downward and steps 101 to 106 are repeated to perform the next iterative calculation.

9. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following steps: Step 100: Obtain the voltage data time series value of each single cell in the automobile battery pack to be diagnosed, and construct a voltage data matrix A; Step 101: Preset the length and width of the initial sliding time window, and use the sliding window to intercept data from the voltage data matrix region by region to form an intercepted voltage matrix B; Step 102: according to the data processing principle, the voltage values ​​in the window are kicked out and reconstructed to form a new voltage matrix D; Step 103: Calculate the Shannon entropy value of each single cell using an optimized entropy value algorithm based on the reconstructed data; Step 104: converting the Shannon entropy value of each monomer based on the optimized Z-score method to obtain a monomer voltage abnormal fluctuation assessment coefficient; Step 105: Identify abnormal monomers based on the security assessment strategy; Step 106: Determine whether there is an abnormal monomer, if yes, execute step 107, otherwise execute step 108; Step 107: If there is an abnormal value, locate the position and occurrence time of the abnormal monomer, further determine the degree of abnormality, and send a corresponding prompt to the driver; Step 108: If there is no outlier, the time window moves downward and steps 101 to 106 are repeated to perform the next iterative calculation.

Images