Electric vehicle battery after-sales analysis method

Abstract

The invention relates to the technical field of after-sales management, in particular to an electric vehicle battery after-sales analysis method, which comprises the following steps of: analyzing a battery cell temperature peak value of a single electric vehicle battery and synchronous charging and discharging current fluctuation, judging whether temperature and current abrupt change are synchronous or not, screening an abnormal time period, normalizing a voltage and current curve and then calculating a node offset proportion, and extracting a working condition characteristic amplitude, performing layered screening on a main control parameter group, and outputting periodic processing suggestion information. According to the method, distribution quantitative analysis is carried out in combination with node differences of periodic voltage and current tracks, collection and deviation extraction of multiple groups of working condition features are realized, main control factors are screened after parameters are layered, targeted suggestion contents are output for periodic state adaptation, the initiative of abnormal working condition identification and the comprehensiveness of periodic change trend analysis can be improved, and the accuracy of abnormal working condition identification is improved. The fault evolution path and working condition distribution are carefully tracked, a suggestion set suitable for periodic working condition difference is formed, and the timeliness of maintenance response and the matching degree of suggestions are enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of after-sales management, and particularly relates to a method for analyzing after-sales of an electric vehicle battery. BACKGROUND

[0002] After-sales management involves systematic management and services carried out for user feedback, product maintenance, fault handling, accessory replacement, performance detection, data collection and analysis, maintenance recommendations, etc. in the whole life cycle of products after sales, including establishing a product traceability system, perfecting a customer service response mechanism, formulating maintenance specifications, improving problem response efficiency, optimizing service quality and improving user experience, etc. Among them, the traditional electric vehicle battery after-sales analysis method refers to, in the process of using an electric vehicle, analyzing the actual working conditions and health status of the battery through artificial detection, battery voltage testing, current detection, charge and discharge performance detection, fault record registration, user feedback collection, after-sales detection report filling, etc. to solve problems such as performance degradation, abnormal failure and recycling detection of the battery in the after-sales link, and then to realize maintenance, replacement or recycling of the battery.

[0003] The prior art mainly develops through artificial detection, battery single parameter testing and working condition registration link. The detection process is mainly static sampling and periodic external recording, which is difficult to capture the working condition changes in the dynamic evolution process. The key parameter changes are only concerned when they are abnormally increased. The working condition fluctuation in the period and the abnormal response period are difficult to accurately depict, which leads to the problems of missed judgment and delayed response in the scenes of unclear battery state boundary, sparse data collection distribution and insensitive abnormal interval discrimination, thereby affecting the rationality of the collaborative diagnosis between parameters and reducing the pertinence and timeliness of the maintenance recommendations. SUMMARY

[0004] The purpose of the present application is to solve the problems existing in the prior art and to provide a method for analyzing after-sales of an electric vehicle battery.

[0005] In order to achieve the above purpose, the present application adopts the following technical scheme: a method for analyzing after-sales of an electric vehicle battery, comprising the following steps: S1: based on the electric vehicle battery, analyzing the temperature peak value of the single battery cell and the fluctuation amplitude of the synchronous charge and discharge current, judging the temperature and current mutation relationship at the collection time point, identifying the parameter synchronous mutation period, and obtaining the cooperative response period interval; S2: based on the cooperative response period interval, judging the working condition data in the period, comparing the voltage curve and the current trajectory after normalization, counting the continuous deviation nodes, and analyzing according to the distribution proportion to obtain the period trajectory deviation proportion; S3: based on the periodic trajectory offset proportion, screening periodic working conditions, interval analysis cell voltage fluctuation, calculating the maximum deviation, judging the SOC fluctuation, identifying the maximum discharge section, obtaining the working condition difference characteristic amplitude; S4: based on the working condition difference characteristic amplitude, comparing the influence of each working condition parameter on the maintenance suggestion, sorting and screening the front parameter group, determining the first layer parameter, counting the number of first layer parameters, obtaining the number of main control parameter groups; S5: based on the number of main control parameter groups, retrieving corresponding maintenance suggestion actions, analyzing the relationship between periodic working conditions and suggestion actions, combining actual working condition parameters, constructing periodic processing content, and obtaining periodic processing suggestion information.

[0006] The application improves that the cooperative response time interval includes start time, end time and abnormal identifier, the periodic trajectory offset proportion includes offset node proportion, periodic coverage rate and abnormal distribution feature, the working condition difference characteristic amplitude includes amplitude distribution type, maximum deviation position and abnormal frequency, the number of main control parameter groups includes parameter name, hierarchical level and priority order, and the periodic processing suggestion information includes suggestion type, operation key points and applicable conditions.

[0007] The application improves that the cooperative response time interval is obtained by the following steps: S111: based on the electric vehicle battery, analyzing the temperature and current fluctuation data of each monomer cell at each collection time point, judging whether the mutation action is consistent for the temperature peak value and the charge and discharge current fluctuation amplitude at each time, comparing the synchronous change trend of temperature and current fluctuation, and obtaining the temperature and current fluctuation joint sequence; S112: based on the temperature and current fluctuation joint sequence, judging whether the change direction of temperature peak value and current fluctuation amplitude at the collection time point is consistent, if there is consistency, screening the synchronous mutation time interval that meets the condition, and obtaining the mutation consistency time interval; S113: based on the mutation consistency time interval, combining complete collection data, analyzing the start and end time points of each synchronous mutation time interval, and labeling them, and obtaining the cooperative response time interval.

[0008] The application improves that the periodic trajectory offset proportion is obtained by the following steps: S211: based on the cooperative response time interval, extracting the data node of the voltage curve in the period, collecting each sampling node of the current trajectory, synchronizing the original sequence data from various sources to the same period dimension, dividing and normalizing the nodes according to the same step length, unifying the number of nodes and recombining the time sequence data, and obtaining the normalized node sequence; S212: Based on the normalized node sequence, the voltage data and the current data of each normalized node are identified in turn, the offset state of the node is compared with each node of the standard baseline curve point by point, the continuous change of adjacent nodes in the voltage and current dimensions is recorded, and a node offset data set is obtained; S213: Based on the node offset data set, the distribution density of the continuous offset paragraph on the periodic normalized timeline is analyzed, the proportion of the paragraph on the timeline is aggregated with the node coverage, the continuous coverage range and the distribution position of each paragraph are recorded, and a periodic trajectory offset proportion is obtained.

[0009] The application improves that the acquisition step of the working condition difference characteristic amplitude is specifically: S311: Based on the periodic trajectory offset proportion, the voltage difference of the single battery cell in the working condition cycle is calculated, all node voltage differences are compared and the node with the largest difference is determined, and then the induction is completed in combination with the trajectory offset proportion, and a maximum voltage deviation is obtained. S312: Based on the maximum voltage deviation, the fluctuation amplitude of the SOC curve is judged, the continuous discharge section is screened in combination with the node sequence of the discharge interval of the working condition cycle, and the maximum deviation is associated and compared, and the number of discharge continuous sections is obtained. S313: Based on the number of discharge continuous sections, the voltage amplitude of the SOC curve fluctuation mapping and the voltage offset of the temperature active interval mapping are calculated, and the working condition difference characteristic amplitude is obtained.

[0010] The application improves that the acquisition step of the number of master parameter groups is specifically: S411: Based on the working condition difference characteristic amplitude, the amplitude distribution type, the maximum deviation position and the abnormal frequency associated with the amplitude are analyzed, the synchronous change of the parameter items in the cycle is identified, it is judged that which parameters are more concentrated in the synchronous fluctuation in the key working condition section, the parameters with correlation are screened, and the working condition parameter coverage is obtained. S412: Based on the working condition parameter coverage, the coverage segments and the paragraphs appearing synchronously in the working condition trajectory of each parameter are compared, the coverage frequency of each parameter in the difference working condition segment is calculated, the parameter set with the coverage frequency greater than the preset coverage threshold is judged, and the number of periodic priority parameter groups is obtained. S413: Based on the number of periodic priority parameter groups, the parameter set at the top of the sorting is screened, the synchronous fluctuation performance of each parameter in the abnormal segment is analyzed, the parameter layering with continuous synchronous performance and change amplitude exceeding the preset fluctuation threshold is optimized, and the number of parameter items at the top of the layering is counted, and the number of master parameter groups is obtained.

[0011] The application improves that the acquisition step of the periodic processing suggestion information is specifically: S511: Based on the number of parameter groups, the parameter name, hierarchical level and priority order are classified, the corresponding maintenance suggestion action is retrieved for each parameter group, the mapping relationship between the parameters and the suggestion action is screened, the combination structure of the suggestion action is judged, and the suggestion action matching information is obtained; S512: Based on the suggestion action matching information, the applicable range of the actual working condition parameters and the suggestion action in the period is analyzed, the fitting degree of the parameter state and the suggestion action adaptation condition is compared, the suggestion action screening strategy is adjusted, and the applicability of each suggestion action in the current period is judged, and the suggestion action adaptation data is obtained; S513: Based on the suggestion action adaptation data, the archiving relationship between the suggestion action and the actual working condition parameters is analyzed, the suggestion action collection method is optimized, the operation points and suggestion types of the suggestion actions with an adaptation score higher than the adaptation threshold are screened, and the period processing suggestion information is obtained.

[0012] The temperature peak value refers to the highest temperature reading of the single cell in a certain period, the working condition data in the period refers to the full amount of collected data of each operating parameter of the battery in a defined period, and the interval analysis refers to the statistical comparison of the voltage variation range of the single cell to identify the fluctuation range in a period.

[0013] Compared with the prior art, the advantages and positive effects of the present application are: In the present application, by dynamically detecting the synchronous change of the key operating parameters, the abnormal working condition section is determined by using the mutation relationship between the cell temperature and the current signal, and the distribution quantitative analysis is carried out by combining the node difference of the periodic voltage and current trajectory, the collection and deviation extraction of multiple working condition characteristics are realized, the parameters are layered and the master control factor is screened, the periodic state adaptation outputs targeted suggestion content, which can improve the initiative of abnormal working condition recognition and the comprehensiveness of periodic change trend analysis, realize the detailed tracking of fault evolution path and working condition distribution, form a suggestion set suitable for the difference of periodic working condition, enhance the timeliness of maintenance response and the matching degree of suggestion, and build a working condition characteristic grading and suggestion generation logic based on data driving. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 The main step flow chart of the present application; Figure 2 The acquisition flow chart of the cooperative response period interval in the present application; Figure 3 The acquisition flow chart of the periodic trajectory offset proportion in the present application; Figure 4 The acquisition flow chart of the working condition difference characteristic amplitude in the present application; Figure 5 The acquisition flow chart of the number of master parameter groups in the present application; Figure 6 The flow chart of the acquisition of the periodic treatment suggestion information in the present application. DETAILED DESCRIPTION

[0015] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not used to limit the present application.

[0016] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0017] EMBODIMENT Please refer to Figure 1 The present application provides a technical solution: a method for analyzing electric vehicle battery after-sales, comprising the following steps: S1: Based on the electric vehicle battery, the temperature peak of the single battery cell and the fluctuation amplitude of the current of the cell at the same period are analyzed, and it is judged whether the change of the temperature peak at each collection point is consistent with the sudden change of the current fluctuation. By synchronously checking the temperature data and the current data, the time period of synchronous mutation of the two parameters is screened, and the cooperative response time interval is obtained; S2: Based on the cooperative response time interval, the working condition data in the period is judged, the voltage curve and the current trajectory in the period are unified according to the normalized node length, the data point difference of each group of nodes and the standard baseline curve is calculated in turn, the total number of nodes that deviate continuously is counted, and the proportion analysis is performed through the node distribution ratio, and the trajectory deviation proportion of the period is obtained; S3: Based on the trajectory deviation proportion of the period, the period working condition is screened, the voltage fluctuation of the single battery cell is analyzed, the maximum deviation of the voltage between the battery cells is calculated, the fluctuation amplitude of the SOC curve in the same period is judged, the number of continuous segments of the maximum discharge interval is identified, the number of charge and discharge in the period and the temperature active interval are determined, and the working condition difference characteristic amplitude is obtained; S4: Based on the working condition difference characteristic amplitude, the influence of each working condition parameter on the maintenance suggestion is compared, the influence degree of each parameter is sorted, the parameter group with a high influence is screened, the parameters belonging to the first layer are determined according to the influence hierarchical structure, the actual number of the first layer parameters is counted, and the number of the main control parameter group is obtained; S5: Based on the number of master parameter groups, retrieve the maintenance recommendation actions corresponding to the parameters, analyze the corresponding relationship between the periodic working conditions and the recommendation actions, combine each recommendation action with the actual working condition parameters, construct the processing content corresponding to the period, and obtain the periodic processing recommendation information.

[0018] The cooperative response period interval includes a start time, an end time, and an abnormality identifier. The periodic trajectory offset proportion includes an offset node proportion, a periodic coverage rate, and an abnormality distribution feature. The working condition difference feature amplitude includes an amplitude distribution type, a maximum deviation position, and an abnormal frequency. The master parameter group number includes a parameter name, a hierarchical level, and a priority order. The periodic processing recommendation information includes a recommendation type, an operation point, and an applicable condition.

[0019] In S1, the temperature peak refers to the highest temperature reading of a single cell at each collection time in a certain period, which reflects the thermal limit performance of the cell under a specific operating state. The charge and discharge current fluctuation amplitude refers to the absolute size of the current change per unit time during charging and discharging, which reflects the dynamic change level of the cell's load bearing and energy release. The collection time point refers to the specific time recording point when the data collection system monitors the cell state in real time, usually every recording time under the collection frequency setting. The sudden action consistency refers to the fact that at the same collection time point, the temperature peak and the current fluctuation amplitude both show obvious rising or falling, etc. rapid changes, and the change direction and timing of the two are consistent. The synchronous check refers to comparing and cross-verifying the change curves of the temperature peak and the charge and discharge current fluctuation amplitude according to the same collection time point, and finding the situation of simultaneous abnormal fluctuation of the two parameters. The period segment of synchronous mutation refers to the data time period in which both the temperature and the current show synchronous and sudden changes. The segment is extracted for subsequent abnormal analysis.

[0020] In S2, the in-cycle working condition data refers to the full collection of data for each operating parameter of the battery within a defined cycle (such as a single complete charging and discharging cycle), including voltage, current, temperature, SOC, etc.; the normalized node length refers to the standardization processing of all curves according to uniform time intervals and data point quantities to ensure fair comparison between different cycles or different data groups; the standard baseline curve refers to an ideal voltage-current change curve under the battery health state constructed according to experience samples, which is used as a reference for comparison of the deviation of each cycle curve; the data point difference calculation refers to the difference (such as absolute difference or deviation) calculation of the measured voltage / current value and the standard baseline value at each node (at the same normalized collection time point); the continuously deviated node refers to a node that is counted as a continuously deviated node if adjacent multiple nodes have obvious deviation in the same direction from the baseline curve in a group of data sequences; the node distribution ratio refers to the ratio of the number of continuously deviated nodes to the total number of data nodes in the current cycle, which is used to quantify the severity of the cycle anomaly; the proportion analysis refers to the analysis of the node distribution ratio and other statistical results to distinguish the influence range and intensity of the deviation anomaly in different cycles.

[0021] In S3, the interval analysis refers to the statistics and comparison of the voltage change range of the single battery cell, identifying the fluctuation range (maximum and minimum values, distribution characteristics, etc.) within a cycle; the maximum deviation refers to the maximum voltage difference between all single battery cells at the same time, reflecting the consistency problem between the single cells, which is an important indicator for judging the balance of battery performance; the fluctuation amplitude in the same cycle refers to the maximum change from the beginning to the end of the cycle in the SOC (state of charge) or other working condition parameter curve, which is used to measure the battery energy utilization and decay trend; the number of continuous segments in the maximum discharge interval refers to the number of time segments during which the single battery cell is continuously in a high discharge state during the discharge process, which is used to reveal the load capacity and stress concentration of the battery cell; the temperature active interval refers to the time interval during which the temperature of the battery cell is in a state of severe fluctuation or high level, reflecting the heat management pressure and abnormal distribution characteristics.

[0022] In S4, each working condition parameter refers to each parameter collected by the battery management system, such as single voltage, SOC, maximum discharge current, cycle number, temperature interval, etc.; the maintenance suggestion refers to the corresponding battery maintenance or care suggestion for different parameter abnormal states, such as replacement, repair, deep detection or limited use prompt; the front-affected parameter group refers to the group of working condition parameters that have the greatest impact after sorting the correlation between each parameter and the maintenance suggestion; the impact hierarchical structure refers to the hierarchical structure divided according to the importance of each parameter to the decision; the first layer parameter refers to the parameter item that belongs to the first level of the impact hierarchical structure and has the most reference value for the maintenance decision in the sorting result.

[0023] In S5, the maintenance suggestion action is proposed for the detected specific parameter exception, such as adjusting the charging and discharging strategy, suggesting replacing the battery, etc.; the corresponding relationship is the logical or rule mapping between the parameters in the period and the maintenance suggestion action, and each parameter exception corresponds to a specific maintenance suggestion; and the actual working condition parameter is the real monitored running data parameter of the battery in the current period, which is directly used as the basis for generating the maintenance suggestion.

[0024] Referring to Figure 2 The acquisition step of the cooperative response time interval is specifically as follows: S111: Based on the battery of the electric vehicle, the temperature and current fluctuation data of the single battery cell at each collection time point are analyzed, and for the temperature peak value and the fluctuation amplitude of the charging and discharging current at each time, it is judged whether the mutation action is consistent, and the temperature and current fluctuation joint sequence is obtained by comparing the synchronous change trend of the temperature and current fluctuation.

[0025] Firstly, the temperature and current fluctuation data of the battery cell at each time point are extracted in the vehicle collection system. For the collection frequency of every second or every five seconds, the temperature reading of the battery cell at the current time and the current difference value of the current and the previous time are read from the historical record to identify the temperature peak value and the fluctuation amplitude of the current. In the set 60-second window, the temperature value at the current time point is compared with all the data in the window to determine whether it is the highest value in the window, thereby confirming the temperature peak value. At the same time, the current value at the current time point and the current value at the previous time point are extracted, and the absolute difference value is calculated to determine the current fluctuation degree. The difference value is compared with the mutation threshold value set according to the sample statistical result. If the difference value exceeds the threshold value, it means that the current fluctuation has mutated. On this basis, it is further judged whether the temperature value at the current time and the temperature change direction at the previous time are consistent with the current change direction, i.e., whether the temperature and the current rise or fall at the same time. For example, at the 120th second, the temperature rises from 44.7 degrees of the previous second to 45.2 degrees, and the current rises from 26.2 amperes of the previous second to 30.0 amperes. The change directions of both are positive, which satisfies the consistency condition of the change trend. At the same time, the current fluctuation is 3.8 amperes, which exceeds the threshold value 3.5 amperes obtained from the previous sample statistics. Therefore, it can be judged that the temperature and the current at this time point constitute an effective synchronous mutation action, and the result is recorded as an event point. The same operation is continued for the subsequent time points to construct a series of time point sequences that satisfy the mutation threshold value and the consistency condition of the change direction, and the time, the temperature of the battery cell, the current difference value and other attributes of each event are recorded to form the temperature and current fluctuation joint sequence.

[0026] S112: Based on the temperature and current fluctuation joint sequence, it is judged whether the change direction of the temperature peak value and the fluctuation amplitude of the current at the collection time point is consistent. If there is consistency, the synchronous mutation time interval that meets the condition is selected to obtain the mutation consistency time interval.

[0027] After obtaining the complete temperature and current fluctuation joint sequence, it is necessary to further identify the time period with consistent change direction. The method is to extract the temperature change direction and current change direction of each event point from the joint sequence one by one, judge whether it remains consistent, that is, the change direction at the continuous time points is all positive or all negative, and the current change amplitude of the event point all exceeds the mutation threshold set by the sample statistics. For the continuous time period that meets the conditions, extract its starting time and ending time as the preliminary synchronous mutation period. In order to prevent the real mutation period from being fragmented due to slight discontinuity of data sampling, the time interval of adjacent two mutation periods is judged. If the interval between the two periods is less than the maximum allowed interval of 30 seconds, it is considered as the same mutation continuous area, and the merging processing is carried out. For example, the first period is from 100 seconds to 105 seconds, the second period is from 120 seconds to 124 seconds, the interval between the two periods is 15 seconds, which is less than the set maximum interval of 30 seconds, so the two periods are merged into a complete synchronous mutation period from 100 seconds to 124 seconds. Repeat the process to screen all period segments that meet the requirements. Each segment meets the preset rules in time continuity, direction consistency and mutation amplitude consistency. The mutation consistency period interval identified is summarized.

[0028] S113: Based on the mutation consistency period interval, the starting and ending time points of each synchronous mutation period are analyzed and labeled based on the complete collected data, and the collaborative response period interval is obtained. After identifying all the mutation consistency periods, the original full collection data needs to be aligned and labeled. The method is to locate the starting time and ending time of each mutation consistency interval from the original data record, extract all original sampling point data in the time interval, add an abnormal identification field to each data record, and mark the time point as a point where synchronous mutation occurs. At the same time, a flag bit is added to each original record to mark whether it is in an abnormal state. If it is in the synchronous mutation interval, it is set to 1, otherwise it is set to 0. The structured data labeling process is realized. For example, it is identified that the period from 120 seconds to 130 seconds is a mutation consistency interval. The identification bit of each record in the 11 sampling points is set to 1, and the mutation type field is added. The whole collection data has the structured attribute of whether it belongs to abnormal mutation on each record. After repeating the process for multiple periods, the accurate alignment and complete labeling of the original data and the mutation period are completed, and the collaborative response period interval with clear structure and perfect field is formed.

[0029] Please refer to Figure 3 The acquisition steps of the periodic trajectory offset proportion are as follows: S211: Based on the cooperative response period interval, extract the data nodes of the voltage curve in the period, collect each sampling node of the current trajectory, synchronize the original sequence data of each source to the same period dimension, divide and normalize the nodes according to the uniform step, unify the number of nodes and reorganize the time sequence data, and obtain the normalized node sequence; First, locate the start and end sampling time in the period, and extract all voltage and current sampling data in the corresponding time period in the original data record. The processing process of the voltage data is to divide each time point voltage reading into the voltage curve node of the period according to the time sequence, ensure that the time interval is uniform and the sampling frequency is consistent, for example, if the cycle length is 600 seconds and the sampling frequency is 1 Hz, it should contain 600 voltage nodes, and the current trajectory needs to be processed synchronously, extract the current value of the same period from the complete current sampling sequence, and perform uniform dimension synchronization operation on the two groups of original sampling node data, that is, map the voltage and current sampling data to the same time line in the same period. Then, determine the uniform node step, for example, set it to the period length divided by the fixed number of nodes, such as normalized to 100 nodes, each node represents a 6-second interval, and the data collection operation is performed in the interval, and a plurality of sampling points in each 6 seconds are aggregated into a representative value. The setting of the node value can be completed by taking the interval average or median. Continue to perform the same division and aggregation operation on the voltage and current data respectively to ensure that the total number of nodes of the two is equal and time-aligned. Then, recombine the normalized voltage sequence and current sequence, arrange them in ascending order of time to form a complete normalized node sequence. In a specific example, if 1200 points are obtained by sampling in a period, they can be compressed into 100 normalized nodes, each node voltage and current is an aggregation result, and is attached with a corresponding period label, to obtain a normalized node sequence with uniform step, consistent node number and time sequence attribute.

[0030] S212: Based on the normalized node sequence, identify the voltage data and current data of each normalized node in turn, compare each node with the standard baseline curve, record the continuous change of adjacent nodes in the voltage and current dimensions, and obtain the node offset data set. The voltage value and current value in each node are read one by one, and are compared with the data of the same node position in the standard baseline point by point. In actual implementation, the standard baseline is generated by a healthy battery under the same working condition and is normalized as a comparison reference. The comparison operation adopts point-to-point absolute difference judgment to extract the voltage offset and current offset of each node, and to judge whether the offset exceeds the maximum allowable offset range set by the reference. The range can be set as a fixed value through the fluctuation statistics of historical healthy data, for example, the voltage offset tolerance is set to 0.05 volts, and the current offset tolerance is set to 1.5 amperes. If the offset of a node exceeds the tolerance, the node is marked as an offset node, and the offset trend is continuously judged to see whether it remains consistent in the adjacent nodes, that is, the voltage and current are continuously increased or continuously decreased. After two or more nodes are identified to meet the trend, it is confirmed that there is a continuous change paragraph. On this basis, the position, length and offset direction of the continuous offset paragraph are recorded as part of the node offset data. For example, in the 10th to 18th node, if the voltage slowly rises from 3.68V to 3.74V, and the current rises from 22.1A to 26.3A, and the offset of each node exceeds the tolerance and fluctuates in the same direction, it is confirmed that the nine nodes constitute a continuous offset paragraph with consistent trend, and are included in the node offset data set. Repeat the process until all node positions are traversed to form a complete node offset data set.

[0031] S213: Based on the node offset data set, analyze the distribution density of the continuous offset paragraph on the periodic normalization timeline, aggregate the proportion of the paragraph on the timeline and the node coverage, record the continuous coverage range and distribution position of each paragraph, and obtain the periodic trajectory offset proportion.

[0032] The continuous offset paragraph refers to the nodes that are continuously consistent in the offset trend (direction, change continuity) from the node offset data set. The nodes are connected in series to form a paragraph by paragraph interval, that is, the nodes with the same offset trend in the node offset data set are combined into a paragraph; Based on each segment of the node offset data set, the starting and ending positions of each segment on the normalized timeline are recorded to determine the distribution density. When performing this operation, first, the index positions of the starting point and the ending point of each offset segment in the normalized sequence are extracted, and the proportion of the segment length to the total number of normalized nodes is calculated. For example, if the normalized sequence has 100 nodes, and a segment of the offset segment covers nodes 30 to 45, the time coverage rate is 16%. Continue to accumulate the coverage proportion of all offset segments to obtain the overall time proportion, and count the distribution position of all offset segments on the timeline. The sequence can be divided into multiple time blocks, such as five equal parts, each with 20 nodes. The proportion of offset segments in each segment is calculated to reflect the distribution trend of the offset. If the offset segments are mainly concentrated in the early and late periods, the distribution is skewed. If they are evenly distributed in all segments, the distribution density is balanced. In addition, it is also necessary to record whether there is overlap, discontinuity or aggregation between the offset segments, and to describe the specific behavior pattern of the offset on the timeline according to the above information. For example, if four offset segments are identified, located at nodes 10-18, 25-30, 58-65 and 80-92, with coverage proportions of 9%, 6%, 8% and 13% respectively, and the cumulative coverage ratio is 36%, and the distribution is in four time segments, then the complete period trajectory offset proportion information can be constructed according to the above information.

[0033] Please refer to Figure 4 The acquisition step of the working condition difference feature amplitude is specifically: S311: Based on the period trajectory offset proportion, calculate the voltage difference of the single battery cell in the working condition period, compare all node voltage differences and determine the node with the largest difference, and then combine the trajectory offset proportion to complete the induction to obtain the maximum voltage deviation. The voltage values of all single batteries on each sampling node in the cycle are extracted, and a voltage matrix arranged in time sequence is constructed. Each row in the voltage matrix represents a time point, and each column represents a battery. For each time point, the voltage values of all batteries in the row are called to calculate the maximum difference between the single batteries at that time. This difference is the voltage difference at that time point. Then, the voltage differences of all time points in the entire cycle are traversed, and the values of each time point are compared to find the time node with the largest voltage difference and record the node position as the region with the most concentrated voltage difference. In this process, an example can be given. For example, at the 240th second in the cycle, the voltage of battery 1 is 3.68V, the voltage of battery 2 is 3.71V, the voltage of battery 3 is 3.73V, and the voltages of the remaining batteries are between 3.69V and 3.70V. The maximum voltage difference at this time point is 3.73V minus 3.68V, which is 0.05V. This value is compared with the voltage differences of other nodes in the cycle. If it is the maximum value, then the node is the target node. Then, while analyzing the maximum voltage deviation, the cycle trajectory offset ratio calculated in the previous step is combined to compare the time sequence of the period with the maximum voltage deviation and the period with the maximum voltage difference. If they occur in similar time periods, it is determined that the maximum voltage deviation and the trajectory offset are highly coincident. Further, the coincident relationship is summarized and arranged to record the maximum voltage deviation time point, battery number, corresponding voltage value, and the number and proportion of overlapping nodes in the overlapping offset interval. Finally, the complete record content is obtained to get the maximum voltage deviation.

[0034] S312: Based on the maximum voltage deviation, the fluctuation amplitude of the SOC curve is determined, the continuous discharge section is selected in combination with the discharge interval node sequence of the working condition cycle, and the number of continuous discharge sections is obtained by associating and comparing the maximum deviation. read the SOC curve data adjacent to its occurrence node, and calculate the fluctuation amplitude of the SOC in the cycle, in the operation process, first extract the starting value and ending value of the SOC curve, calculate the difference between the two, then analyze the overall change trajectory of the curve in the cycle, judge whether there are multiple fluctuation intervals, judge whether the discharge process is continuous according to the SOC change curvature, if the curve continuously decreases in a certain length range and the slope value is in the stable change range, it is determined that this time period is the discharge section, on this basis, read the discharge interval node sequence in the cycle, extract each node paragraph marked as discharge state, and select the paragraph whose continuous decline time exceeds the set threshold, for example, set the continuous decline duration to more than 30 seconds, then if a discharge state maintains 40 seconds, it is included in the effective paragraph, otherwise if it only maintains 15 seconds, it is not counted, then, call the obtained voltage maximum deviation value, extract the voltage difference of the middle or tail node of each continuous discharge section, and compare the value with the maximum deviation, judge whether the deviation presents a synchronous fluctuation trend in the current discharge section, if the voltage difference in a discharge section continuously approaches or exceeds 80% of the reference value of the maximum deviation, it is determined that the discharge section has significant correlation with the deviation, and it is counted into the number of effective discharge duration sections, for example, six discharge sections are identified in the cycle, four of which have voltage deviation values exceeding 80% of the maximum deviation, it is confirmed that the cycle has four effective discharge duration sections, combine the node sequence and deviation data to get the number of discharge duration sections.

[0035] S313: based on the number of discharge duration sections, calculate the voltage amplitude of the SOC curve fluctuation mapping and the voltage offset of the temperature active interval mapping, using the formula: ; get the working condition difference characteristic amplitude , represent the total number of nodes in the cycle, represent the the maximum voltage deviation of the node, which is the voltage difference of the node relative to other single cells, represent the the voltage amplitude offset obtained by the SOC curve fluctuation of the node, reflecting the voltage difference corresponding to the state of charge change, represent the the number of discharge duration sections, representing the number of continuous discharge sections formed in the node cycle, represent the the voltage offset obtained by the temperature active interval mapping of the node, reflecting the influence of temperature change on voltage distribution.

[0036] The working condition difference characteristic amplitude is a "total difference amount" formed by the voltage difference, SOC fluctuation and temperature influence of the battery monomer in one charging and discharging cycle, and reflects the abnormal strength degree of the battery operation working condition in the cycle.

[0037] The SOC curve fluctuation and temperature active interval are structured and processed, different participants are unified to the voltage offset dimension, first, the maximum deviation amount of the voltage of each node in the cycle is extracted and normalized according to the voltage reference interval, and the original physical meaning is retained, for example, the original maximum voltage deviation of node 1 is 12.5mV, and after normalization, it is represented as , the original maximum voltage deviation of node 2 is 9.8mV, and after normalization, it is represented as , the original maximum voltage deviation of node 3 is 11.3mV, and after normalization, it is represented as , then the fluctuation amplitude of the SOC curve is extracted, the SOC difference between nodes is fitted and transformed according to the voltage response, and the SOC voltage mapping offset of each node is obtained, node 1 is 4.7mV after mapping, and is normalized as , node 2 is 3.9mV, and is normalized as , node 3 is 5.1mV, and is normalized as , and the number of discharge segments is taken as the intensity parameter, the temperature active interval is quantified, the nodes with significant changes in the temperature curve are extracted, and the disturbance amplitude is linearly mapped to form a voltage offset representation; The temperature fluctuation influence of node 1 is 2.6mV, and is normalized as ; The temperature fluctuation influence of node 2 is 3.1mV, and is normalized as ; The temperature fluctuation influence of node 3 is 2.8mV, and is normalized as ; The number of discharge duration segments corresponding to each node is , , , the total number of nodes is set to , and the formula is used for calculation: ; Substitute the normalized parameter values: , , , , ; , , , , ; , , , , ; Calculate each item in turn: First item: ; Second item: ; Third item: ; Substitute the formula and perform the sum and square root operations: ; According to the normalized analysis results of the historical data distribution characteristics in the implementation process, the judgment criteria for the working condition difference characteristic amplitude are divided into the following three continuous intervals: When , it is determined to be a "normal interval", indicating that there is no cooperative amplification behavior between the voltage deviation, SOC fluctuation and temperature disturbance in the period, and the battery is in a stable running state; When , it is determined to be a "transition interval", indicating that there is a slight abnormal correlation fluctuation in the period, which needs to be conditionally monitored but not directly maintained; When , it is determined to be an "abnormal interval", indicating that there is a significant parameter cooperative deviation in the period, which needs to be marked as an abnormal period and trigger the subsequent main control parameter identification process.

[0038] The results show that the working condition difference characteristic amplitude obtained in the current period has entered the "abnormal interval", indicating that there is a significant cooperative amplification relationship between the voltage difference, SOC voltage mapping deviation and temperature disturbance voltage response in the period, and the current period behavior belongs to an abnormal characteristic trajectory; The formula maps the voltage difference, SOC disturbance and temperature response to the same physical dimension and integrates the dynamic discharge behavior in a normalized and weighted manner, thereby establishing a correlation between different types of abnormalities, which helps to identify the abnormal characteristic amplitude in the working condition period.

[0039] Please refer to Figure 5 , the acquisition steps of the number of main control parameter groups are as follows: S411: Based on the working condition difference characteristic amplitude, analyze its amplitude distribution type, maximum deviation position and abnormal frequency, identify the synchronous change of parameters in the period, judge which parameters fluctuate more concentratedly in the key working condition section, select parameters with correlation, and obtain the working condition parameter coverage; First, the time series data of all working condition parameters is extracted from the identified abnormal paragraphs in the cycle, and each parameter is labeled in the cycle. For each parameter labeled as abnormal, the fluctuation amplitude of each parameter in different time periods in the cycle is counted, the frequency distribution of the fluctuation amplitude in all fluctuation data is calculated, and the distribution interval is divided according to the amplitude value, for example, the amplitude can be divided into a stable interval below 0.01, a moderate fluctuation interval between 0.01 and 0.03, and a mutation interval above 0.03, so as to determine the amplitude distribution type of the parameter. Subsequently, each parameter's fluctuation data is traversed and searched, and the position of the most severe value change in the time series of the parameter is located to identify the maximum deviation position of the parameter in the cycle. For example, the cell temperature rises from 35.4°C to 38.8°C at 300 seconds, which is the maximum change position in the cycle. Then the number of times the parameter enters the abnormal interval in the cycle is counted, which is the abnormal frequency. For example, if the single cell voltage fluctuates more than 0.04 volts five times in the cycle, the abnormal frequency is 5 times. After the above statistics are completed for each parameter, a record item containing three attributes is formed, including amplitude distribution type, maximum deviation position, and abnormal frequency. Then, according to the time label, it is determined whether multiple parameters fluctuate in the same period. If the current, voltage, SOC and temperature all change more than the set mutation threshold in a certain working condition, and the time difference is not more than 2 seconds, it is considered that the parameters in this period have synchronous change performance. The parameter items participating in synchronous change are recorded, and it is determined whether the synchronous fluctuation behavior is concentrated in some key working condition period, such as the discharge peak period or the temperature active period. If the number of synchronous parameter fluctuations in this working condition period accounts for more than 60%, it is considered as a concentrated paragraph of synchronous mutation, and the parameter list appearing in this paragraph is considered as a parameter group with correlation. Then the number of parameters in each paragraph is counted to obtain the working condition parameter coverage.

[0040] S412: Based on the working condition parameter coverage, compare the coverage segments of each parameter in the working condition trajectory with the paragraphs that appear synchronously, calculate the coverage frequency of each parameter in the different working condition segments, and determine the parameter set whose coverage frequency is greater than the preset coverage threshold to obtain the number of cycle priority parameter groups. The time slice of the fluctuation occurrence of each parameter item participating in the synchronous mutation is extracted, and is compared with each difference working condition segment divided in the cycle. First, the start and end time of the abnormal paragraph of each parameter in the cycle is determined, then the corresponding difference segment list is called in the working condition track, whether the parameter fluctuation time overlaps with the difference working condition time is compared, if the overlapping interval length exceeds 10 seconds, it is considered that the parameter covers the working condition segment, the cumulative of this kind of covering behavior is calculated, the covering frequency of the parameter in all difference segments is calculated, the frequency value is equal to the number of covering segments divided by the total number of difference segments, for example, 12 difference segments are identified in the cycle, if SOC parameter appears in 9 of them, the covering frequency of SOC is 75%, the same operation is performed on each parameter item, and the covering frequency is sorted, it is judged which parameter covering frequency reaches the outstanding level, in the numerical setting, the covering frequency greater than or equal to 70% can be used as the parameter screening basis of the covering performance, all parameter items with frequency value not less than the threshold value can be included in the cycle priority parameter set, then the number of parameters in the set is accumulated, and the total number of parameters participating in the covering and showing outstanding performance is recorded as the number of priority parameter groups in the cycle.

[0041] S413: Based on the number of priority parameter groups in the cycle, the parameter set with high ranking is screened and sorted, the synchronous fluctuation performance of each parameter in the abnormal segment is analyzed, the parameters with continuous and changed amplitude exceeding the preset fluctuation threshold are layered, and the number of the first parameter items is counted to obtain the number of master parameter groups. All parameter sets with high ranking after sorting are summarized, in the set, the specific performance of each parameter in the abnormal segment in the cycle is analyzed, the fluctuation value sequence of each abnormal segment is extracted, and it is judged whether the fluctuation change is continuous and concentrated. In this judgment, if the proportion of the number of continuous fluctuation nodes to the total number of nodes of the abnormal segment exceeds 60%, and the fluctuation direction is continuously the same, the change amplitude remains in the rising or falling trend, the parameter is determined to have good synchronous performance. Continue to count all parameters with this feature, and divide them into layers according to the fluctuation concentration degree and the synchronous duration time, for example, the parameters with continuous synchronous fluctuation time exceeding 30 seconds and fluctuation amplitude standard deviation less than 0.02 are set as the first level, the rest are divided into the second level and below. The actual number of parameters in the first level is summarized as the number of master parameter groups in the cycle. For example, if there are 10 parameters in the priority parameter group, 4 of which show synchronous continuous fluctuation, concentrated change and reach the first level standard in multiple abnormal segments, the number of master parameter groups in the cycle is output as 4.

[0042] Please refer to Figure 6 , the acquisition step of the cycle processing suggestion information is specifically: S511: Based on the number of master parameter groups, the parameter name, hierarchical level and priority order are classified, the corresponding maintenance suggestion action is retrieved for each parameter group, the mapping relationship between the parameter and the suggestion action is screened, the combination structure of the suggestion action is judged, and the suggestion action matching information is obtained; Extract the parameter name, corresponding hierarchical level and priority order of all master parameters, and arrange them in ascending order according to the priority, construct a structured parameter list, and each parameter item in the list needs to be attached with complete attribute value, for example, the parameter "single body voltage" belongs to the first hierarchical level and the priority is 1, the parameter "SOC" belongs to the second hierarchical level and the priority is 2, and so on. Then, the retrieval process of maintenance suggestion action is performed for each parameter item, the keyword index or attribute matching is found through the preset maintenance suggestion database, the corresponding maintenance action of the parameter in the historical case or operation and maintenance rule is located, for example, for voltage abnormality, the retrieval result is "limit maximum charging current" or "suggest replacing module", and such suggestion action is associated with the corresponding parameter one by one to form a preliminary mapping relationship table, in which each parameter corresponds to one or more suggestion actions. Then, the aggregation classification operation is performed on the mapping relationship, whether the same type of parameter corresponds to multiple suggestion actions and the coverage frequency are analyzed, if a parameter is associated with three suggestion actions and the remaining parameters are associated with only one suggestion, the parameter has multiple suggestion structure, and then the suggestion action combination structure of the parameter group is judged, whether there are repeated actions, conflicting actions or dependent action relationship is checked, if multiple parameters are associated with the same maintenance suggestion action, it is marked as repeated relationship, if two parameter suggestion actions are logically exclusive, for example, "suggest replacing module" and "continue running observation", it is marked as conflict relationship, all relationships are summarized and the corresponding parameter combination structure is recorded, and the suggestion action matching information is obtained.

[0043] S512: Based on the suggestion action matching information, the application range of the actual working condition parameters and the suggestion action in the period is analyzed, the fitting degree of the parameter state and the suggestion action adaptation condition is compared, the selection strategy of the suggestion action is adjusted, and the applicability of each suggestion action in the current period is judged, and the suggestion action adaptation data is obtained; The matching degree of the actual collected working condition parameters in the current period and the corresponding recommended action in the set applicable range is analyzed item by item, the applicable condition field of each recommended action is extracted, for example, "voltage fluctuation exceeds 0.08 volts" or "SOC drop rate is continuously higher than 2% per minute", and then it is searched whether there is a time segment that meets the above conditions in the actual working condition data of the current period. If the working condition parameters reach the starting threshold set by the recommended action in any time period, it is recorded as a successful match, and if it does not reach at all, it is recorded as a mismatch. Then the matching of the recommended action and the parameter is compared one by one. If most of the sampling points in the period meet the recommended action conditions, it is determined that the matching degree is high, and if only individual edge periods meet the conditions, the matching degree is low. On this basis, the recommended action is re-screened and sorted, and the high matching degree action is placed in the priority list. The matching degree of the action is not clear or the edge matching action is further adjusted and the judgment rule is adjusted, for example, a 5% buffer zone is set for the boundary value, that is, the recommended action requires that the current fluctuation exceeds 10 amperes, and the actual fluctuation is 9.6 amperes, which is considered to approximately meet the conditions and is included in the matching reference range. The applicability of all recommended actions is summarized and counted, the state matching, starting proportion and adaptation state of each recommended action in the current period are recorded, and the recommended action adaptation data is obtained.

[0044] S513: Based on the recommended action adaptation data, the archival relationship between the recommended action and the actual working condition parameter is analyzed, the collection mode of the recommended action is optimized, the operation points and recommended types of the recommended actions with adaptation degree scores higher than the adaptation threshold are screened, and the period processing recommendation information is obtained; Firstly, the archival relationship between all the successfully adapted recommended actions and their corresponding parameter value intervals is established. In the archival operation process, the applicable condition range of each recommended action, the corresponding actual period matched and the working condition parameters involved are bound into a unified record item. In the data structure, the fields of recommended action, triggering parameter, parameter value range, actual starting time period and matching degree level are included. Then, the archived recommended actions are classified and collected according to similar triggering conditions, for example, multiple recommended actions are triggered based on voltage mutation, or multiple actions are triggered by high temperature interval, and then they are classified into the same collection category. It is further judged whether there is a recommended action group with higher matching strength in each collection category. The starting number of recommended actions in each group and the average of adaptation level are counted, and the recommended actions with starting number more than twice and adaptation level of "high" are selected and marked. The operation point content in the action, such as "current limiting", "deep detection" or "replacement of the whole package", is extracted as the recommended operation direction. At the same time, the recommended types are screened and classified, such as "conservative monitoring type", "moderate intervention type", "immediate maintenance type" and the like. The selected recommended action points and their recommended types are collected and output, and the period processing recommendation information is obtained.

[0045] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any skilled person in the art can modify or change the disclosed technical content into equivalent embodiments with equivalent changes, and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution content of the present application, still falls within the protection scope of the present application.

Claims

1. A method for after-sales analysis of electric vehicle batteries, characterized in that, The method comprises the following steps: S1: based on the battery of the electric vehicle, analyzing the temperature peak value of the single battery cell and the fluctuation amplitude of the current during the same period, judging the temperature and current mutation relationship at the collection time point, identifying the parameter synchronous mutation period, and obtaining the cooperative response period interval; S2: based on the cooperative response period interval, judging the working condition data in the period, comparing the voltage curve and the current trajectory after normalization, counting the continuous deviation nodes, and analyzing the distribution proportion to obtain the period trajectory deviation proportion; S3: based on the period trajectory deviation proportion, screening the period working condition, analyzing the battery cell voltage fluctuation, calculating the maximum deviation, judging the SOC fluctuation, identifying the maximum discharge section, and obtaining the working condition difference characteristic amplitude; S4: based on the working condition difference characteristic amplitude, comparing the influence of each working condition parameter on the maintenance suggestion, sorting and screening the front parameter group, determining the first layer parameter, counting the number of first layer parameters, and obtaining the number of main control parameter groups; S5: based on the number of main control parameter groups, retrieving the corresponding maintenance suggestion action, analyzing the relationship between the period working condition and the suggestion action, combining the actual working condition parameters, constructing the period processing content, and obtaining the period processing suggestion information.

2. The method for post-sale analysis of an electric vehicle battery of claim 1, wherein, The cooperative response period interval includes a start time, an end time, and an abnormality identifier. The period trajectory deviation proportion includes a deviation node proportion, a period coverage rate, and an abnormality distribution characteristic. The working condition difference characteristic amplitude includes an amplitude distribution type, a maximum deviation position, and an abnormality frequency. The number of main control parameter groups includes a parameter name, a hierarchical level, and a priority order. The period processing suggestion information includes a suggestion type, an operation key point, and an applicable condition.

3. The method for post-sale analysis of an electric vehicle battery of claim 1, wherein, The acquisition step of the cooperative response period interval is specifically: S111: based on the battery of the electric vehicle, analyzing the temperature and current fluctuation data of the single battery cell at each collection time point, judging whether the mutation action is consistent for the temperature peak value and the charge-discharge current fluctuation amplitude at each time, comparing the synchronous change trend of the temperature and current fluctuation to obtain a temperature and current fluctuation joint sequence; S112: based on the temperature and current fluctuation joint sequence, judging whether the change direction of the temperature peak value and the current fluctuation amplitude at the collection time point is consistent, if there is consistency, screening the synchronous mutation period that meets the condition, and obtaining the mutation consistency period interval; S113: based on the mutation consistency period interval, combining the complete collection data, analyzing the start and end time points of each synchronous mutation period, and labeling them to obtain the cooperative response period interval.

4. The method for post-sale analysis of electric vehicle batteries of claim 1, wherein, The acquisition step of the period trajectory deviation proportion is specifically: S211: based on the cooperative response period interval, extracting the data nodes of the voltage curve in the period, collecting each sampling node of the current trajectory, synchronizing the original sequence data from various sources to the same period dimension, dividing and normalizing the nodes according to the same step, unifying the number of nodes and reorganizing the time sequence data, and obtaining the normalized node sequence; S212: Based on the normalized node sequence, the voltage data and the current data of each normalized node are identified in turn, the offset state of the node is compared with each node of the standard baseline curve point by point, the continuous change of adjacent nodes in the voltage and current dimensions is recorded, and a node offset data set is obtained; S213: Based on the node offset data set, the distribution density of the continuous offset paragraph on the period normalized timeline is analyzed, the proportion of the paragraph on the timeline and the node coverage are aggregated, the continuous coverage range and the distribution position of each paragraph are recorded, and a period trajectory offset proportion is obtained.

5. The method for post-sale analysis of electric vehicle batteries of claim 1, wherein, The acquisition step of the working condition difference characteristic amplitude is specifically: S311: Based on the period trajectory offset proportion, the voltage difference of the single battery cell in the working condition period is calculated, all node voltage differences are compared and the node with the largest difference is determined, and the induction is completed in combination with the trajectory offset proportion to obtain the maximum voltage deviation; S312: Based on the maximum voltage deviation, the fluctuation amplitude of the SOC curve is judged, the continuous discharge section is screened in combination with the discharge interval node sequence of the working condition period, and the maximum deviation is associated and compared to obtain the number of continuous discharge sections; S313: Based on the number of continuous discharge sections, the voltage amplitude of the SOC curve fluctuation mapping and the voltage offset of the temperature active interval mapping are calculated to obtain the working condition difference characteristic amplitude.

6. The method for post-sale analysis of electric vehicle batteries of claim 1, wherein, The acquisition step of the number of main control parameter groups is specifically: S411: Based on the working condition difference characteristic amplitude, the amplitude distribution type, the maximum deviation position and the abnormal frequency associated with it are analyzed, the synchronous change of the parameter items in the period is identified, it is judged which parameters fluctuate synchronously in the key working condition section, the parameters with correlation are screened, and the working condition parameter coverage is obtained; S412: Based on the working condition parameter coverage, the coverage segment and the paragraph appearing synchronously in the working condition trajectory of each parameter are compared, the coverage frequency of each parameter in the difference working condition segment is calculated, the parameter set with the coverage frequency greater than the preset coverage threshold is judged, and the number of period priority parameter groups is obtained; S413: Based on the number of period priority parameter groups, the parameter set at the top of the sorting is screened, the synchronous fluctuation performance of each parameter in the abnormal segment is analyzed, the parameter layering with continuous synchronous performance and change amplitude exceeding the preset fluctuation threshold is optimized, and the number of parameter items at the top of the layering is counted to obtain the number of main control parameter groups.

7. The method for post-sale analysis of electric vehicle batteries of claim 1, wherein, The acquisition step of the period processing suggestion information is specifically: S511: Based on the number of main control parameter groups, the parameter name, layering level and priority order are classified, the corresponding maintenance suggestion action of each parameter group is retrieved, the mapping relationship between the parameter and the suggestion action is screened, the combination structure of the suggestion action is judged, and the suggestion action matching information is obtained; S512: Based on the suggestion action matching information, the application range of the actual working condition parameter and the suggestion action in the period is analyzed, the fitting degree of the parameter state and the suggestion action adaptation condition is compared, the selection strategy of the suggestion action is adjusted, and the applicability of each suggestion action in the current period is judged to obtain suggestion action adaptation data; S513: Based on the recommended action adaptation data, analyze the archival relationship between the recommended action and the actual working condition parameters, optimize the collection method of the recommended action, and filter the operation points and recommended types of the recommended action with an adaptation score higher than the adaptation threshold to obtain periodic processing recommendation information.

8. The method for post-sale analysis of electric vehicle batteries of claim 1, wherein, The temperature peak value refers to the highest temperature reading of the single battery cell at each collection time in a certain period. The working condition data in the period refers to the full collection data of each operating parameter of the battery in a defined period. The interval analysis refers to the statistics and comparison of the voltage variation range of the single battery cell to identify its fluctuation range in a period.

Images