CN-122020031-A - Electric automobile battery comprehensive evaluation method and system based on charging pile

CN122020031ACN 122020031 ACN122020031 ACN 122020031ACN-122020031-A

Abstract

The invention discloses a comprehensive evaluation method and a system for an electric vehicle battery based on a charging pile, and relates to the field of comprehensive evaluation of electric vehicle batteries, wherein the method comprises the steps of acquiring vehicle core operation data in real time based on a standard charging protocol, and constructing an automobile charging data set; the method comprises the steps of establishing a multi-dimensional feature vector for battery state evaluation by extracting key physical behavior features of a charging process according to an automobile charging data set, establishing a vehicle battery function evaluation model according to the multi-dimensional feature vector for battery state evaluation, outputting the health state of an electric automobile battery in the charging process, and automatically generating a digital report comprising battery health evaluation, safety risk evaluation and function state evaluation according to the output result of the vehicle battery function evaluation model. The method has the advantages that the method realizes the normalized and automatic comprehensive evaluation of the health state, the safety risk and the system function of the electric automobile battery by collecting the charging data in real time and extracting the multidimensional characteristics.

Inventors

Fu Rongcheng
FAN JIEHUI
ZHANG XINGHUI
ZHANG JIAQING
ZHENG YUJIA

Assignees

广州车数新能源科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260120

Claims (10)

1. The electric automobile battery comprehensive evaluation method based on the charging pile is characterized by comprising the following steps of: the method comprises the steps that a charging pile deployed in a wide area is used as a distributed data acquisition terminal, vehicle core operation data are acquired in real time based on a standard charging protocol, and an automobile charging data set is constructed; according to the automobile charging data set, a multidimensional feature vector for battery state evaluation is constructed by extracting key physical behavior features in the charging process; according to the multidimensional feature vector for battery state evaluation, a vehicle battery function evaluation model is constructed, and the state of health of the electric vehicle battery in the charging process is output; And automatically generating a digital report comprising the battery health evaluation, the safety risk evaluation and the functional state evaluation according to the output result of the vehicle battery functional evaluation model.
2. The method for comprehensively evaluating the battery of the electric vehicle based on the charging pile according to claim 1, wherein the method for acquiring the vehicle core operation data in real time based on the standard charging protocol by using the widely deployed charging pile as the distributed data acquisition terminal, and constructing the vehicle charging data set specifically comprises the following steps: When the vehicle is connected with the charging pile for charging, CAN message data in GB/T27930 protocol is captured in real time through the charging pile controller; acquiring a BMS handshake message and a BMS identification message, and extracting vehicle static data in a handshake stage; Analyzing BCS, BCP and BSM messages in real time, and extracting dynamic high-frequency data in a charging stage; And constructing an automobile charging data set according to the vehicle static data and the dynamic high frequency data.
3. The method for comprehensively evaluating the battery of the electric automobile based on the charging pile according to claim 2, wherein the constructing the multidimensional feature vector for evaluating the state of the battery by extracting key physical behavior features of the charging process according to the automobile charging dataset specifically comprises: Acquiring charging data of each complete charging session in an automobile charging data set, and preprocessing; Three core features representing a charging behavior mode, a battery core consistency state and thermal management efficiency are extracted from the preprocessed charging dynamic time sequence data, and the method specifically comprises the following steps: Based on different stages of the charging process, key change rate characteristics of voltage and current are calculated respectively, and charging acceptance characteristics of the battery are quantized; calculating a dispersion index of the battery cell voltage difference according to the distribution characteristic of the change of the charge state, and quantitatively evaluating the consistency of the battery cells in the battery pack; Calculating the temperature rise rate under a specific working condition through a response curve of the change of the battery temperature along with the charging power, and evaluating the real-time efficiency of the battery thermal management system; Based on the key change rate characteristics, the dispersion index and the temperature rise rate of the voltage and the current, a multidimensional characteristic vector for battery state evaluation is constructed.
4. The method for comprehensively evaluating the battery of the electric automobile based on the charging pile according to claim 3, wherein the constructing a vehicle battery function evaluation model according to the multidimensional feature vector for battery state evaluation, and outputting the health state of the battery of the electric automobile specifically comprises: Building a vehicle battery function evaluation model, wherein the vehicle battery function evaluation model takes a multidimensional feature vector as input and takes a health state in the charging process of an electric vehicle battery as output; the model adopts a parallelization submodule architecture and comprises a battery health degree evaluation module, a safety risk evaluation module and a functional state evaluation module, wherein each module outputs an evaluation result based on a multidimensional feature vector; The battery health evaluation module adopts an ampere-hour integration method and a direct current internal resistance measurement method to cross-verify and quantitatively calculate the health state index of the battery from two dimensions of energy storage capacity and charge transfer efficiency; the safety risk assessment module establishes an abnormality detection mechanism based on real-time data flow, identifies time sequence mutation of key safety parameters and compares the time sequence mutation with a threshold value, and establishes hierarchical early warning of thermal runaway risk and cell consistency degradation; the functional state evaluation module evaluates the running state and reliability of the vehicle battery management system and the charging control related functions through carrying out statistical analysis on the message interaction quality, protocol compliance and abnormal termination reasons in the charging communication process.
5. The method for comprehensively evaluating the battery of the electric automobile based on the charging pile according to claim 4, wherein the battery health evaluation module specifically comprises: Screening complete charging fragments in the charging process of the same charging pile based on 3 sigma criteria according to historical data of charging of batteries of the electric automobile; based on an ampere-hour integration method, performing time integration on the current in the charging segment, and calculating the total electric quantity actually charged; estimating the actual chargeable capacity corresponding to the current charge by combining the SOC value at the start and stop time of the charge; calculating a capacity retention rate according to the estimated actual chargeable capacity and the rated total capacity of the battery in the vehicle static data; in the charging process, identifying the moment when the current is subjected to step change, wherein if the absolute value of the current change quantity of two adjacent sampling points exceeds a set value, judging that the step change occurs; Acquiring the variation of the voltage in a short time before and after the current step, and estimating the direct-current internal resistance in the charging process; And calculating the internal resistance increase proportion according to the calculated direct current internal resistance value and the standard internal resistance value of the same type of battery in the healthy state.
6. The method for comprehensively evaluating the battery of the electric vehicle based on the charging pile according to claim 4, wherein the safety risk evaluation module specifically comprises: Establishing a hierarchical threshold trigger based on temperature rise rate time sequence data of a maximum temperature point of the battery; by continuously monitoring a temperature-power response curve and combining a grading threshold value and duration, the state of charging power is utilized for cross verification, and normal fast charging heat generation and abnormal internal short circuit or heat dissipation faults are distinguished; And calculating the difference value of the highest single voltage and the lowest single voltage in the battery pack in real time, judging that the voltage imbalance exists among the cells when the average value of a plurality of continuous detection periods of the voltage difference exceeds a safety threshold value, and generating a cell consistency risk early warning.
7. The method for comprehensively evaluating the battery of the electric vehicle based on the charging pile according to claim 4, wherein the functional state evaluation module specifically comprises: based on BMS periodic message types and sending frequency agreed by GB/T27930 protocol, determining the total number of periodic messages which the BMS should send in the whole charging session; Counting the packet loss rate of the periodic messages by comparing the total number of the periodic messages with the number of the messages actually received by the charging pile; presetting a BMS response timeout threshold a3, and recording the times that the BMS response time exceeds the threshold after the charging pile sends a control instruction; Analyzing a CST message sent by the BMS when the charging is stopped, extracting a code of 'stopping the charging reason' in the CST message, and converting the code into a numerical code by adopting a single-hot code; Training a decision tree model based on the history labeling data, taking the packet loss rate, the response timeout times and the pause reason code as input characteristics, and outputting a classification result of the functional state of the vehicle charging system; Wherein, the classification result comprises three grades of perfect function, slight abnormal function and defective function.
8. The method for comprehensively evaluating the battery of the electric vehicle based on the charging pile according to claim 7, wherein the automatically generating the digital report including the battery health evaluation, the safety risk evaluation and the functional status evaluation according to the output result of the vehicle battery functional evaluation model specifically comprises: Based on the unified charging session identifier, the output result of each evaluation module, the vehicle static data, the dynamic time sequence data and the charging process statistical data are related to the same data object; Invoking a predefined and structured report template, wherein the template defines placeholders and presentation rules of all data fields, and automatically filling the associated data objects to corresponding positions of the template according to field mapping; Based on the completed template content, a final visual report document or structured data package is generated by a rendering engine.
9. The electric automobile battery comprehensive evaluation system based on the charging pile is characterized by being used for realizing the electric automobile battery comprehensive evaluation method based on the charging pile as claimed in any one of claims 1-8, and comprising the following steps: The data acquisition module is used for acquiring vehicle core operation data in real time based on a standard charging protocol by using the widely deployed charging piles as distributed data acquisition terminals and constructing an automobile charging data set; The comprehensive evaluation module is used for constructing a multi-dimensional feature vector for battery state evaluation by extracting key physical behavior features of a charging process according to an automobile charging data set; And the visualization module is used for automatically generating a digital report comprising battery health evaluation, safety risk evaluation and functional state evaluation according to the output result of the vehicle battery functional evaluation model.
10. The electric vehicle battery comprehensive assessment system based on the charging pile according to claim 9, wherein the comprehensive assessment module comprises: The characteristic vector unit is used for constructing a multidimensional characteristic vector for battery state evaluation by extracting key physical behavior characteristics of a charging process according to an automobile charging data set; And the comprehensive evaluation unit is used for constructing a vehicle battery function evaluation model according to the multidimensional feature vector for battery state evaluation and outputting the health state of the electric vehicle battery in the charging process.

Description

Electric automobile battery comprehensive evaluation method and system based on charging pile Technical Field The invention relates to the field of comprehensive evaluation of batteries of electric vehicles, in particular to a comprehensive evaluation method and system of batteries of electric vehicles based on charging piles. Background Along with the rapid development of new energy automobile industry, the health Status (SOH), safety performance and functional reliability of the power battery are directly related to the running safety, endurance and service life of the vehicle, and become key factors influencing the vehicle purchasing decision of users, the transaction estimation of second-hand vehicles and the sustainable development of industry, and currently, the accurate assessment requirements of new energy automobile users, operation and maintenance institutions, vehicle enterprises and supervision departments on the state of the power battery are urgent, especially in a charging scene, the battery is in a high-load and high-risk running stage, and real-time and comprehensive state monitoring and assessment means are more needed. The charging pile is used as an indispensable supporting infrastructure of a new energy automobile, wide-area deployment is realized, and the charging pile can directly interact with a vehicle Battery Management System (BMS) in a protocol in a charging process, so that the charging pile has the natural advantage of acquiring static basic information and dynamic operation data of a battery. However, the core function of the existing charging pile is still limited to energy transmission, mass data generated in the charging process cannot be fully utilized to carry out deep analysis of battery states, so that data resources are wasted, the core requirements of industry on full life cycle management and safety risk early warning of batteries cannot be met, under the background, a distributed data acquisition and evaluation system is constructed by means of the wide-area charging pile, comprehensive state evaluation of power batteries is realized, and the method becomes an important technical direction for solving industry pain points. At present, aiming at the health degree (SOH) and safety assessment of a power battery of a new energy automobile, the method mainly relies on the following modes of self-checking of a vehicle-mounted BMS (battery management system), namely, depending on a vehicle algorithm, the data is generally black-boxed, a user or a third party is difficult to acquire real original data, the BMS estimation precision of an old automobile model is lower, and the method requires professional personnel to read by using specific equipment for disassembly or OBD equipment detection, has complex operation and high cost, cannot be carried out at high frequency, and requires the transportation of the vehicle or a battery pack to a laboratory for charge and discharge test, is time-consuming and labor-consuming, has no popularity, so that the following problems are caused: The data displayed by the vehicle is often beautified by the vehicle enterprise, but the transaction of the second hand vehicle lacks public trust; Most of the existing charging piles are only used as 'energy transmission pipelines', and abnormal data (such as overlarge voltage difference and abnormal temperature rise) in the charging process are not subjected to deep excavation and real-time safety evaluation; The failure of the static data (vehicle base information) to combine effectively with the dynamic data (charging real-time data) results in the assessment model lacking multi-dimensional support and "data islanding" occurring. Disclosure of Invention In order to solve the technical problems, the technical scheme provides an electric vehicle battery comprehensive evaluation method and system based on a charging pile, which solves the problems that data displayed by a vehicle in the background technology is often beautified by a vehicle enterprise, public trust is lacked in a second-hand vehicle transaction, a plurality of battery thermal runaway occurs in charging or after charging, the existing charging pile is mainly only used as an 'energy transmission pipeline', abnormal data in the charging process is not subjected to deep mining and real-time safety evaluation, static data and dynamic data are not effectively combined, so that an evaluation model lacks multi-dimensional support, and a 'data island' problem occurs. In order to achieve the above purpose, the invention adopts the following technical scheme: An electric automobile battery comprehensive evaluation method based on a charging pile comprises the following steps: the method comprises the steps that a charging pile deployed in a wide area is used as a distributed data acquisition terminal, vehicle core operation data are acquired in real time based on a standard charging protocol, and an automobile charging data set is constructed;