Battery model parameter online identification and updating method and system based on cloud collaboration

Abstract

The invention provides a battery model parameter online identification and updating method and system based on cloud collaboration, and relates to the technical field of ecological risk assess.The method comprises the steps that battery operation data of a to-be-tested vehicle is collected, and the initial charge state corresponding to the to-be-tested vehicle is calculated according to the battery operation data; constructing an operation database in a cloud server according to all battery operation data and the initial charge state, and screening from the operation database based on a preset screening condition to obtain an effective data segment; based on a GA-LM joint optimization algorithm and the effective data segments, performing optimization identification on identification parameters of the second-order RC battery model and outputting an identification parameter group; and performing parameter optimization on the vehicle end battery management system according to the identification parameters in the identification parameter group, and outputting a predicted battery state. The prediction reliability of the state of charge and the state of health of the battery can be improved.

Description

Technical Field

[0001] This invention relates to the field of ecological risk assessment technology, and in particular to a method and system for online identification and updating of battery model parameters based on cloud collaboration. Background Technology

[0002] Climate change, the energy crisis, and environmental pollution are three major problems facing human society today, all of which are inextricably linked to the consumption of petroleum resources and emissions from gasoline-powered vehicles. With the development of electric vehicle technology and the increasing environmental awareness of governments worldwide, more and more countries are beginning to regard the development of electric vehicle technology as an important measure to alleviate environmental problems.

[0003] Chinese patent CN114035072B discloses a cloud-edge collaborative method for joint estimation of battery pack states of charge (SOC) and health status at different time scales. The method involves periodic information interaction between a cloud control platform and an onboard battery management system to jointly estimate SOC and health status of the vehicle's power battery pack at different time scales. The onboard terminal uses a real-time estimation framework, integrating online parameter identification, adaptive state filtering, ampere-hour integration, open-circuit voltage static correction, full-charge correction, and maximum / minimum voltage cell selection strategies to achieve accurate and robust online SOC estimation. The cloud control platform utilizes machine learning algorithms to periodically update the SOC / health status of the battery pack. The onboard terminal updates the SOC of the battery pack in real time and periodically uploads it along with current, voltage, and temperature signals to a big data center for model training and prediction. The cloud control platform periodically updates the SOC and health status of the battery pack and sends it to the onboard terminal to calibrate the SOC estimation. However, the above scheme lacks a valid data filtering mechanism for different operating conditions in the data processing flow and fails to consider key factors such as sufficiency of current excitation and temperature stability. This results in the model training using a large amount of low-quality data, reducing the accuracy of parameter identification and the reliability of state estimation. Therefore, it is essential to provide a cloud-based collaborative method and system for online identification and updating of battery model parameters to improve the reliability of predicting state of charge and battery health. Summary of the Invention

[0004] In view of this, the present invention proposes a method and system for online identification and updating of battery model parameters based on cloud collaboration.

[0005] This invention provides a cloud-based collaborative method for online identification and updating of battery model parameters, the method comprising: Collect battery operation data of the vehicle under test, and calculate the initial state of charge of the vehicle under test based on the battery operation data; An operational database is built on the cloud server based on all battery operating data and the initial state of charge, and valid data segments are obtained by filtering from the operational database based on preset filtering conditions. Based on the GA-LM joint optimization algorithm and the effective data segment, the identification parameters of the second-order RC battery model are optimized and identified, and the identification parameter set is output. The vehicle-side battery management system is optimized based on the identification parameters in the identification parameter group, and a predicted battery state is output.

[0006] Based on the above technical solutions, preferably, the preset screening conditions include current excitation sufficiency conditions, temperature stability conditions, SOC variation range screening conditions, and data continuity conditions. Specifically, the current excitation sufficiency condition is that the standard deviation of the current within the effective data segment is greater than a preset multiple threshold of the battery capacity; the temperature stability condition is that the temperature fluctuation value within the effective data segment is less than or equal to a preset temperature threshold; the SOC variation range screening condition is that a first slope interval is selected for lithium iron phosphate batteries, and a second slope interval is selected for ternary lithium batteries, with the first slope interval and the second slope interval partially overlapping; and the data continuity condition is that the time period length corresponding to the effective data segment is greater than or equal to a preset time length.

[0007] Based on the above technical solutions, preferably, the optimization and identification of the identification parameters of the second-order RC battery model based on the GA-LM joint optimization algorithm and the effective data segment, and the output of the identification parameter set, specifically includes: The original parameters of the GA-LM joint optimization algorithm are initialized, including gene encoding method, population size, number of iterations, and individual fitness. The fitness function of the GA-LM joint optimization algorithm is determined based on the individual fitness and the number of individuals in the population, and the fitness function is used to evaluate each individual in the population, and the individuals in the population are sorted according to their fitness. A crossover operation is performed on random individuals in the population to generate offspring individuals, and a mutation operation is performed on each offspring individual to generate new genomes and update the population. If the number of iterations of the GA-LM joint optimization algorithm reaches the preset number of iterations or the fitness of the current individual is greater than the preset fitness, then the iteration process of the GA-LM joint optimization algorithm ends, the individual with the highest fitness in the last generation population is selected as the optimal solution, and the optimization parameters are obtained by decoding to optimize and identify the identification parameters of the second-order RC battery model and output the identification parameter set.

[0008] More preferably, the method further includes: The identification parameter set, the average temperature of the data segment corresponding to the identification parameter set, and the average state of charge corresponding to the identification parameter set are used to construct an identification parameter triplet. The identification parameter triplet is then written into the parameter mapping relationship diagram corresponding to the battery of the vehicle under test. The parameter mapping relationship diagram is a multi-dimensional lookup table, with the input dimensions being temperature and state of charge, and the output dimension being battery model parameters. The parameter mapping relationship diagram provides the vehicle-side battery management system with the most matching parameter values ​​to achieve online adaptive updating of the model.

[0009] More preferably, the step of optimizing the vehicle-side battery management system based on the identification parameters in the identification parameter group and outputting a predicted battery state specifically includes: If the parameter mapping relationship diagram of the vehicle under test meets the update conditions, the cloud server sends the identification parameter group to the vehicle-side battery management system. The update conditions include that the difference between the identification parameter group and the original identification parameter group is greater than a preset identification difference value, or that the time of sending by the cloud server is greater than the time of the last sending than a preset update period.

[0010] More preferably, the expression corresponding to the second-order RC battery model is:

[0011] in, U RC1 ( k ) indicates the first k Electrochemical polarization voltage at time, τ 1 represents the time constant corresponding to the electrochemical polarization circuit. R 1 represents the electrochemical polarization resistance. Indicates the sampling time interval. I ( k ) indicates the first k Battery current at any given moment U RC2 ( k ) indicates the first k Concentration polarization voltage at time, τ 2 represents the time constant corresponding to the concentration polarization loop. R 2 represents the concentration polarization internal resistance. U L ( k ) indicates the first k Battery terminal voltage at time, U oc ( k ) No. k The battery open-circuit voltage at any given time. R 0 represents the ohmic internal resistance of the battery.

[0012] More preferably, the identification parameters include ohmic internal resistance, electrochemical polarization internal resistance, electrochemical polarization capacitance, concentration polarization internal resistance, and concentration polarization capacitance.

[0013] A second aspect of this application provides a cloud-based collaborative online identification and updating system for battery model parameters. This system includes a data acquisition module, a data processing module, and an identification and updating module. The data acquisition module is used to collect battery operation data of the vehicle under test and calculate the initial state of charge of the vehicle under test based on the battery operation data. The data processing module is used to build an operating database on the cloud server based on all battery operating data and the initial state of charge, and to filter from the operating database based on preset filtering conditions to obtain valid data segments; based on the GA-LM joint optimization algorithm and the valid data segments, the identification parameters of the second-order RC battery model are optimized and identified, and the identification parameter set is output. The identification update module is used to optimize the parameters of the vehicle-side battery management system based on the identification parameters in the identification parameter group and output a predicted battery status.

[0014] A third aspect of this application provides an electronic device including a processor, a memory, a user interface, and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory.

[0015] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the steps of a cloud-based collaborative method for online identification and updating of battery model parameters.

[0016] The cloud-based collaborative online identification and updating method and system for battery model parameters provided by this invention has the following advantages over existing technologies: (1) By aggregating data from multiple vehicles and operating conditions on the cloud server and filtering effective segments, the interference of noise and outliers on identification is significantly reduced. The GA-LM joint optimization algorithm has both global search and fast local convergence capabilities, reducing the probability of getting trapped in local optima and shortening the convergence time. Furthermore, it updates model parameters online as the battery ages, temperature and operating conditions change, suppressing parameter drift and improving the predictive reliability of state of charge and battery health throughout the entire life cycle. At the same time, it achieves early warning through parameter and state deviation monitoring, reducing the probability of thermal runaway and failure. Large-scale calculation and optimization are completed in the cloud, while the vehicle mainly performs parameter loading and lightweight inference, reducing hardware computing power and storage requirements. Multi-vehicle collaborative learning reduces the workload and time cost of offline testing and repeated calibration for single vehicles.

[0017] (2) By performing a global search in the GA stage and maintaining the diversity of solutions, the probability of getting trapped in local optima is reduced, providing high-quality initial values ​​for the LM stage, which is conducive to fast and high-precision convergence. Fitness ranking and elite retention accelerate convergence and avoid the loss of excellent solutions. At the same time, fitness is defined and evaluated on the effective data segment to suppress the influence of noise and abnormal working conditions and improve the generalization ability of parameters to multiple working conditions and multiple temperatures. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A schematic diagram of the online identification and updating method for battery model parameters based on cloud collaboration provided by the present invention; Figure 2 The topology diagram of the second-order RC equivalent circuit model provided by this invention; Figure 3 The flowchart for cloud-based collaborative parameter identification and update provided by this invention; Figure 4 A schematic diagram of the structure of the online identification and updating system for battery model parameters provided by the present invention; Figure 5 This is a schematic diagram of the structure of the electronic device provided by the present invention.

[0020] Explanation of reference numerals in the attached diagram: 1. Online identification and update system for battery model parameters; 11. Data acquisition module; 12. Data processing module; 13. Identification and update module; 2. Electronic equipment; 21. Processor; 22. Communication bus; 23. User interface; 24. Network interface; 25. Memory. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] Before introducing the embodiments of the present invention, some terms and their abbreviations involved in the embodiments of the present invention will be defined and explained.

[0023] Genetic algorithms (GA) are based on natural biological evolution. Through selection, crossover, and mutation (genetic operators), they solve optimization problems. After the initial population is randomly generated, the genetic operators iteratively generate better approximate solutions generation by generation. The Levenberg-Marquardt (LM) algorithm is a classic nonlinear optimization algorithm that integrates the advantages of Gauss-Newton's method and gradient descent. This algorithm is very effective in solving problems with redundant parameters and can effectively reduce the possibility of the objective function getting trapped in local optima.

[0024] This invention discloses a method and system for online identification and updating of battery model parameters based on cloud collaboration, with reference to... Figure 1 The steps of this method include S1 to S4.

[0025] Step S1: Collect battery operating data of the vehicle under test, and calculate the initial state of charge of the vehicle under test based on the battery operating data. Identified parameters include ohmic internal resistance, electrochemical polarization internal resistance, electrochemical polarization capacitance, concentration polarization internal resistance, and concentration polarization capacitance.

[0026] In this step, battery operating data is collected in real time by a data acquisition module installed on the vehicle. This data includes, but is not limited to, voltage, current, temperature, and timestamps, and the initial state of charge (initial SOC) is calculated. The battery operating data is then preprocessed and packaged, and periodically uploaded to a cloud server via an in-vehicle T-Box.

[0027] Furthermore, the battery operation data of the vehicle under test is collected through the vehicle's Battery Management System (BMS) and an onboard remote data acquisition terminal (T-Box). The data collection process is carried out at a preset fixed sampling frequency, and the main data collected includes: battery pack terminal voltage, current, individual battery cell temperature or average battery pack temperature, and corresponding sampling timestamps. If necessary, charge / discharge status indicators and vehicle operating mode information can also be recorded simultaneously.

[0028] The collected raw data needs to be preprocessed, including noise filtering to eliminate noise interference generated during high-frequency sampling; identifying and removing outlier data points, such as abrupt changes in voltage or current and missing data points; and performing time synchronization processing on current and voltage sampling to ensure data temporal consistency. After preprocessing, the data is organized and packaged into standard data packets in chronological order to prepare for subsequent uploading and model building.

[0029] Secondly, based on the collected and preprocessed battery operating data, the initial state of charge (SOC) of the vehicle under test is calculated. The calculation process primarily employs a combination of the integral method and the open-circuit voltage correction method. Specifically, the integral method is based on the fundamental definition of battery SOC and calculates the remaining charge by integrating the battery's charging and discharging current in real time. That is, starting from the SOC at a certain initial moment, the ratio of the integral of the discharging current to the rated capacity is subtracted to obtain the current SOC. The SOC value obtained by the integral method and the SOC value obtained by the open-circuit voltage method are then weighted and fused to form a comprehensive initial SOC estimate. After the calculation is completed, the cloud stores the battery voltage, current, temperature, initial SOC, and timestamp as a complete data record, representing the vehicle's initial operating state information.

[0030] Step S2: Build an operating database on the cloud server based on all battery operating data and initial state of charge, and filter the operating database based on preset filtering conditions to obtain valid data segments.

[0031] In this step, the preset screening conditions include current excitation sufficiency, temperature stability, SOC variation range, and data continuity. The current excitation sufficiency condition is that the standard deviation of the current within the effective data segment is greater than a preset multiple threshold of the battery capacity. The temperature stability condition is that the temperature fluctuation value within the effective data segment is less than or equal to a preset temperature threshold. The SOC variation range screening condition is that the first slope interval is selected for lithium iron phosphate batteries, and the second slope interval is selected for ternary lithium batteries. The first slope interval and the second slope interval partially overlap. The data continuity condition is that the time period length corresponding to the effective data segment is greater than or equal to a preset time length.

[0032] In one example, after receiving data packets, the cloud communication interface parses them and stores them in a time-series database or relational database, indexed by VIN code and time. The cloud algorithm service periodically scans the database, filtering valid data segments for each vehicle. The filtering criteria include current excitation sufficiency, temperature stability, SOC variation range, and data continuity. The sufficiency condition for current excitation can be expressed as:

[0033] Where Capacity represents battery capacity, std() represents standard deviation function, and k represents threshold coefficient, that is, the standard deviation of current in the data segment must be greater than 0.2C. For example, for a 60Ah battery, it must be greater than 12A, which can adapt to battery packs of different capacities.

[0034] The scenarios that can be met simultaneously are generally driving discharge, fast charging, and supercharging conditions; and driving discharge needs to be maintained for a certain period of time. Among these, the frequent start-stop in urban conditions may only partially meet the requirements, while the requirements for expressway / highway conditions can generally be met.

[0035] Temperature stability conditions:

[0036] This means the temperature fluctuation range within the data segment is less than or equal to 3°C. Based on real-vehicle data, during high-speed driving, fast charging, and the initial stages of supercharging, the battery generates heat rapidly. Temperature fluctuations are significant when thermal management is not activated, but stabilize after thermal management is activated, with a temperature difference less than or equal to 3°C, which meets most operating conditions. Note: A current standard deviation greater than 0.2C within 10 minutes is sufficient, easily met by most driving conditions. Temperature stability within 10 minutes is relatively easy to achieve. According to the full-vehicle fast-charge test data from the summer calibration experiment, 40%-90% SOC meets the temperature stability requirement. During fast charging, the initial temperature rise is too rapid. After thermal management is activated, the battery temperature tends to stabilize. At the end of fast charging, the battery temperature begins to decrease, with a temperature drop rate exceeding 3°C / 10 minutes, failing to meet the temperature stability requirement. According to the full-vehicle supercharging test data from the summer calibration experiment, 40%-95% SOC meets the temperature stability requirement, but after 95% SOC, the current standard deviation does not meet the 0.2C excitation condition.

[0037] SOC variation range screening criteria: For lithium iron phosphate batteries, the OCV-SOC curve is very flat in the middle section and is very sensitive to the error of the initial SOC value. Therefore, when screening, the flat area should be avoided, and the area with a larger slope at both ends of the SOC should be selected first, that is, the range of [0%, 15%] or [85%, 100%]. For ternary lithium batteries, the slope of the OCV-SOC curve is larger and the plateau area is less. In the middle SOC area (20%-90%), the curve has a continuous and obvious slope, and the voltage and SOC have a good monotonic correspondence.

[0038] Since batteries are generally classified into two types—lithium iron phosphate (LFP) and ternary lithium—they differ significantly in voltage platforms and lifespans. Ternary lithium batteries have lower internal resistance, while LFP batteries have relatively higher internal resistance and significantly different discharge curves. Therefore, they should be processed separately, and data screening criteria should be adjusted accordingly. Current screening criteria are not based on the absolute value of the current, but rather on a ratio relative to the rated capacity, which can accommodate battery packs of different capacities. Usage scenarios are categorized into high-frequency and low-frequency vehicles, and processed separately. The system can also optimize screening and identification strategies based on the characteristics of vehicle usage scenarios: For high-frequency vehicles, due to the large amount of data and sufficient excitation, higher screening standards can be set, such as requiring longer data segments and greater current fluctuations, and performing high-frequency parameter identification daily or every two days. For low-frequency vehicles, such as ordinary private cars, since it is difficult to obtain complete, sufficiently excited long data segments, a "building block" strategy can be adopted: multiple short-distance trips with similar operating conditions, such as similar average temperatures and similar SOC ranges, are intelligently spliced ​​together to form a virtual "long data segment" for parameter identification. Meanwhile, the trigger identification cycle can be extended to weekly or bi-weekly.

[0039] Data continuity condition: The data segment length must last at least 10 minutes, that is, the data segment must contain at least 600 data points and be uninterrupted.

[0040] Step S3: Based on the GA-LM joint optimization algorithm and the effective data segment, optimize and identify the identification parameters of the second-order RC battery model and output the identification parameter set.

[0041] Please see Figure 2 Battery models, due to their good accuracy and simple structure, are widely used equivalent circuit models that characterize lithium batteries using ideal electrical components. Second-order RC equivalent circuit models offer a balance of moderate complexity and high accuracy. Figure 2 This is a schematic diagram of a second-order RC equivalent circuit model.

[0042] Using a discretized second-order RC equivalent circuit model, its state-space equations are as follows:

[0043] in, U RC1 ( k ) indicates the first k The electrochemical polarization voltage at time t, i.e., the voltage across R1 and C1. τ 1 represents the time constant corresponding to the electrochemical polarization circuit. R 1 represents the electrochemical polarization resistance. Indicates the sampling time interval. I ( k ) indicates the first k Battery current at any given momentU RC2 ( k ) indicates the first k The concentration polarization voltage at time t, i.e., the voltage across R2 and C2. τ 2 represents the time constant corresponding to the concentration polarization loop. R 2 represents the concentration polarization internal resistance. U L ( k ) indicates the first k Battery terminal voltage at time, U oc ( k ) No. k The battery open-circuit voltage at any given time. R 0 represents the ohmic internal resistance of the battery.

[0044] There are 5 parameters to be identified. x =[ R 0, R 1, R 2, C 1, C 2]. The fitness function is the reciprocal of the root mean square error (RMSE) between the model-estimated terminal voltage Vest and the measured terminal voltage Vmeas, used as the fitness function of the genetic algorithm to minimize the voltage error. F ( x = 1 / (RMSE(Vmeas,Vest)) Based on the above state equations, for real-vehicle battery data, each current value I and voltage value U has a definite set of parameters to be identified. x =[ R 0, R 1, R 2, C 1, C 2).

[0045] Since the parameters are unknown and only initial values ​​are available, a set of approximate optimal solutions can be obtained through a search algorithm. For a sampling point (current and voltage values) corresponding to a fixed SOC and temperature point, the approximate optimal solution is substituted to obtain the corresponding estimated terminal voltage value. The reciprocal of the root mean square error between the estimated terminal voltage value and the actual vehicle voltage value is used as the objective function for search optimization.

[0046] Please see Figure 3The GA-LM joint optimization algorithm combines the globally convergent features of the GA and LM algorithms. Compared to the traditional GA algorithm, the GA-LM algorithm introduces the LM operator to improve the LM algorithm's shortcomings in local optimization while avoiding the premature convergence phenomenon of the GA algorithm. This algorithm utilizes the initial values ​​of the predicted parameter vectors from the GA algorithm to provide prior information for the LM algorithm. Even without prior values, the GA-LM algorithm can achieve high-precision nonlinear optimization.

[0047] This step also includes steps S31 to S34.

[0048] Step S31: Initialize the original parameters of the input GA-LM joint optimization algorithm. The original parameters include gene encoding method, population size, number of iterations, and individual fitness.

[0049] In this step, during parameter identification of the battery model, due to the lack of initial parameter estimation information, a maximum population size of 100 and 300 iterations are set as stopping criteria to give the algorithm a wider search capability. Real-number encoding is used, which has the advantage of eliminating the need for numerical conversion and allowing direct application of genetic algorithms to the solution's representation, making it easier to combine with classical optimization methods.

[0050] The roulette wheel selection algorithm was used, meaning selection was based on fitness ratios. The probability of an individual being selected was calculated as follows:

[0051] in, F i Indicates the first in the population i The fitness value of each individual N Indicates the number of individuals in the population. F n Indicates the first in the population n The fitness value of each individual.

[0052] Step S32: Determine the fitness function of the GA-LM joint optimization algorithm based on the individual fitness and the number of individuals in the population, and use the fitness function to evaluate each individual in the population, and sort the individuals in the population according to their fitness.

[0053] Step S33: Perform a crossover operation on random individuals in the population to generate offspring individuals, and perform a mutation operation on each offspring individual to generate new genomes and update the population.

[0054] In this step, for the crossover operation, since the individuals use real-number encoding, the real-number crossover method is used. Specifically, the first... kChromosomes a k and the l Chromosomes a l exist j The bit crossover operation method is as follows:

[0055] in, a kj Indicates the first k The chromosome in the _ ... j Gene values ​​after crossover operation. a ij Indicates the first i The chromosome in the _ ... j The original gene value of the position, a lj Indicates the first l The chromosome in the _ ... j The original gene value of the position, b This represents a random number in the interval [0,1].

[0056] Mutation is used to generate superior individuals. An individual is randomly selected from the population, and a specific point within that individual is mutated. i The individual's first j One gene a ij The procedure for performing mutations is as follows:

[0057] in, a max Indicates gene a ij The upper realm, a min Indicates gene a ij The lower bound, r 2 represents a random number in the interval [0,1]. g Indicates the current iteration number. G max Indicates the maximum number of evolutions. f ( g The number represents the mutation probability. The closer the current iteration number is to the maximum number of evolutions, the smaller the mutation probability. The mutation operator refers to randomly changing certain gene values ​​in the individual's encoding string with a certain probability, usually 0.001 to 0.1.

[0058] Step S34: If the number of iterations of the GA-LM joint optimization algorithm reaches the preset number of iterations or the fitness of the current individual is greater than the preset fitness, then the iteration process of the GA-LM joint optimization algorithm ends, the individual with the highest fitness in the last generation population is selected as the optimal solution, and the optimization parameters are obtained by decoding to optimize and identify the identification parameters of the second-order RC battery model and output the identification parameter set.

[0059] In this step, the current individual is calculated. x* ( i fitness function value P* ( x i ),like P* ( x i If the value exceeds the set threshold, continue the mutation operation and record the value of the current best individual. x* ( i ), that is, when P* ( x i )> P ( x i ),but x = x* ( i Conversely, if the genetic algorithm iteration ends, the LM algorithm is directly used for optimization to obtain the optimal individual value. x ( i ).

[0060] Repeat the algorithm following the steps described above. The algorithm terminates when the objective function value reaches a preset value or the algorithm reaches a preset number of iterations. The iteration count is set to 300. When this number is reached, the best individual in the current population is designated as the algorithm's result. Reasonable upper and lower limits are set based on battery characteristics (e.g., R0: [0.5mΩ, 5mΩ]). The optimization process continues iterating until the maximum number of generations is reached or the fitness function converges. The final output is the parameter set X that minimizes the voltage error. opt .

[0061] Furthermore, the identification parameter triplet is constructed by identifying the parameter set, the average temperature of the data segment corresponding to the identification parameter set, and the average state of charge corresponding to the identification parameter set. The identification parameter triplet is then written into the parameter mapping relationship diagram corresponding to the battery of the vehicle under test. The parameter mapping relationship diagram is a multi-dimensional lookup table, with the input dimensions being temperature and state of charge, and the output dimension being the battery model parameters. The parameter mapping relationship diagram provides the most matching parameter values ​​for the vehicle-side battery management system to achieve online adaptive updating of the model.

[0062] After identification, calculate the average temperature T corresponding to this set of parameters. avg and average SOCavg . (T) avg SOC avg ,X opt As a data point, it is updated in the battery's unique parameter mapping diagram (MAP). This MAP is a multi-dimensional lookup table that provides the most suitable battery parameters to the vehicle based on real-time temperature and SOC. The initial MAP can be obtained experimentally and is generated by interpolating a three-dimensional MAP table of SOC, temperature, and battery parameters to assist in SOC estimation and SOH estimation. The updated parameter MAP does not consider SOH. Generally, SOH is calculated directly by looking up the cyclic decay table. After updating the battery parameter MAP, the latest battery parameters are obtained to assist in SOH estimation and correct the SOH lookup value. SOH (aging dimension) can be stored in the cloud and processed according to the same cell type. After accumulating a large amount of data, it can assist in the adaptive updating and verification of the vehicle-side parameter MAP.

[0063] In this embodiment, global search (GA) is used to maintain solution diversity, reduce the probability of getting trapped in local optima, and provide high-quality initial values ​​for the LM stage, which is beneficial for fast and high-precision convergence. Fitness ranking and elite retention accelerate convergence and avoid the loss of excellent solutions. Fitness is defined and evaluated on the effective data segment to suppress the influence of noise and abnormal operating conditions, and improve the generalization ability of parameters to multiple operating conditions and temperatures. The optimized parameters better reproduce polarization and transient characteristics, reduce voltage RMSE / MAE and peak deviation, reduce SOC estimation drift, and improve prediction consistency under rebound voltage and high current conditions. Population individual fitness evaluation can be parallelized, adapting to cloud distributed computing and shortening the total identification time.

[0064] Step S4: Optimize the parameters of the vehicle-side battery management system based on the identification parameters in the identification parameter group and output the predicted battery status.

[0065] In this step, if the parameter mapping relationship diagram of the vehicle under test meets the update conditions, the cloud server sends the identification parameter group to the vehicle battery management system. The update conditions include that the difference between the identification parameter group and the original identification parameter group is greater than the preset identification difference value, or that the time of the cloud server's sending is greater than the time of the last sending than the preset update period.

[0066] Furthermore, when a vehicle's parameter map is significantly updated, for example, if the difference between the new identification point and the original point exceeds 10%, or if more than 7 days have passed since the last distribution, the cloud will trigger a distribution task. For a second-order RC battery model, the battery parameters include ( R 0、 R 1. C 1. R 2. C2) There are five parameters in total, forming five initial MAP tables such as SOC-T-R0. If the value of the MAP point of R0 changes by more than 10% after the update, the latest value is sent out. The same applies to the other battery parameters. The cloud packages the new parameter set or the entire updated parameter MAP table into a message and sends it to the VIN code of the target vehicle through a secure link. After receiving the message, the vehicle's T-Box verifies its integrity and forwards it to the BMS. The BMS will write the new parameters into the non-volatile memory (NVM) before the vehicle is next powered on but in the "Ready" state, replacing the old parameters. After the update is completed, the BMS's subsequent SOC, SOP, and other state estimation algorithms will immediately run with the new parameters, thereby achieving online updating of model parameters and performance improvement.

[0067] In this embodiment, by aggregating multi-vehicle, multi-condition data and filtering effective segments on a cloud server, the interference of noise and outliers on identification is significantly reduced. The GA-LM joint optimization algorithm combines global search and fast local convergence capabilities, reducing the probability of getting trapped in local optima and shortening convergence time. Furthermore, it updates model parameters online as the battery ages, temperature, and operating conditions change, suppressing parameter drift and improving the predictive reliability of state of charge and battery health throughout the entire lifespan. Simultaneously, early warning is achieved through parameter and state deviation monitoring, reducing the probability of thermal runaway and failure. Large-scale computation and optimization are completed in the cloud, while the vehicle mainly performs parameter loading and lightweight inference, reducing hardware computing power and storage requirements. Multi-vehicle collaborative learning reduces the workload and time cost of single-vehicle offline testing and repeated calibration.

[0068] Based on the above method, this application discloses a cloud-based collaborative online battery model parameter identification and update system, referencing... Figure 4 The online identification and update system 1 for battery model parameters includes a data acquisition module 11, a data processing module 12, and an identification and update module 13, wherein... The data acquisition module 11 is used to collect the battery operation data of the vehicle under test and calculate the initial state of charge of the vehicle under test based on the battery operation data. The data processing module 12 is used to build an operating database on the cloud server based on all battery operating data and initial state of charge, and to filter from the operating database based on preset filtering conditions to obtain valid data segments; based on the GA-LM joint optimization algorithm and valid data segments, it optimizes and identifies the identification parameters of the second-order RC battery model and outputs the identification parameter set. The identification update module 13 is used to optimize the parameters of the vehicle-side battery management system based on the identification parameters in the identification parameter group and output the predicted battery status.

[0069] In one example, the preset screening conditions include current excitation sufficiency, temperature stability, SOC variation range, and data continuity. The current excitation sufficiency condition is that the standard deviation of the current within the effective data segment is greater than a preset multiple threshold of the battery capacity. The temperature stability condition is that the temperature fluctuation value within the effective data segment is less than or equal to a preset temperature threshold. The SOC variation range screening condition is that a first slope interval is selected for lithium iron phosphate batteries, and a second slope interval is selected for ternary lithium batteries. The first slope interval and the second slope interval partially overlap. The data continuity condition is that the time period length corresponding to the effective data segment is greater than or equal to a preset time length.

[0070] In one example, the data processing module 12 initializes the original parameters of the input GA-LM joint optimization algorithm, including gene encoding method, population size, number of iterations, and individual fitness. It determines the fitness function of the GA-LM joint optimization algorithm based on individual fitness and population size, and uses this fitness function to evaluate each individual in the population, sorting them by fitness. It performs pairwise crossover on random individuals in the population to generate offspring, and performs mutation on each offspring to generate new genomes and update the population. If the number of iterations of the GA-LM joint optimization algorithm reaches a preset number or the fitness of the current individual is greater than a preset fitness, the iteration process of the GA-LM joint optimization algorithm ends. The individual with the highest fitness in the last generation is selected as the optimal solution, and the optimization parameters are obtained through decoding to optimize and identify the identification parameters of the second-order RC battery model, outputting the identification parameter set.

[0071] In one example, the data processing module 12 is used to construct an identification parameter triplet by taking the identification parameter group, the average temperature of the data segment corresponding to the identification parameter group, and the average state of charge corresponding to the identification parameter group, and write the identification parameter triplet into the parameter mapping relationship diagram corresponding to the battery of the vehicle under test. The parameter mapping relationship diagram is a multi-dimensional lookup table, with the input dimension being temperature and state of charge, and the output dimension being battery model parameters. The parameter mapping relationship diagram provides the most matching parameter values ​​for the vehicle-side battery management system to realize online adaptive updating of the model.

[0072] In one example, the identification update module 13 is used to send an identification parameter group to the vehicle battery management system if the parameter mapping relationship diagram of the vehicle under test meets the update conditions. The update conditions include that the difference between the identification parameter group and the original identification parameter group is greater than a preset identification difference value, or that the time of the cloud server's sending is greater than the time of the last sending than a preset update period.

[0073] In one example, the expression for the second-order RC battery model is:

[0074] in, U RC1 ( k ) indicates the first k Electrochemical polarization voltage at time, τ 1 represents the time constant corresponding to the electrochemical polarization circuit. R 1 represents the electrochemical polarization resistance. Indicates the sampling time interval. I ( k ) indicates the first k Battery current at any given moment U RC2 ( k ) indicates the first k Concentration polarization voltage at time, τ 2 represents the time constant corresponding to the concentration polarization loop. R 2 represents the concentration polarization internal resistance. U L ( k ) indicates the first k Battery terminal voltage at time, U oc ( k ) No. k The battery open-circuit voltage at any given time. R 0 represents the ohmic internal resistance of the battery.

[0075] In one example, the identified parameters include ohmic internal resistance, electrochemical polarization internal resistance, electrochemical polarization capacitance, concentration polarization internal resistance, and concentration polarization capacitance.

[0076] Please see Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 5 As shown, the electronic device 2 may include: at least one processor 21, at least one network interface 24, user interface 23, memory 25, and at least one communication bus 22.

[0077] The communication bus 22 is used to enable communication between these components.

[0078] The user interface 23 may include a display screen and a camera. Optionally, the user interface 23 may also include a standard wired interface and a wireless interface.

[0079] The network interface 24 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0080] The processor 21 may include one or more processing cores. The processor 21 connects to various parts of the server using various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory 25, and by calling data stored in the memory 25. Optionally, the processor 21 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 21 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 21 and may be implemented as a separate chip.

[0081] The memory 25 may include random access memory (RAM) or read-only memory. Optionally, the memory 25 may include non-transitory computer-readable storage medium. The memory 25 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 25 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 25 may also be at least one storage device located remotely from the aforementioned processor 21. Figure 5 As shown, the memory 25, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for an online identification and update method for battery model parameters based on cloud collaboration.

[0082] exist Figure 5In the electronic device 2 shown, the user interface 23 is mainly used to provide an input interface for the user and obtain the user input data; while the processor 21 can be used to call the application stored in the memory 25, which is a cloud-based collaborative method for online identification and updating of battery model parameters. When executed by one or more processors, the electronic device executes one or more methods as described in the above embodiments.

[0083] A computer-readable storage medium storing instructions that, when executed by one or more processors, cause a computer to perform one or more methods as described in the embodiments above.

[0084] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0085] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0086] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings or direct couplings or communication connections may be through some service interfaces; indirect couplings or communication connections between apparatuses or units may be electrical or other forms.

[0087] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0088] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0089] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0090] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for online identification and updating of battery model parameters based on cloud collaboration, characterized in that, The method includes: Collect battery operation data of the vehicle under test, and calculate the initial state of charge of the vehicle under test based on the battery operation data; An operational database is built on the cloud server based on all battery operating data and the initial state of charge, and valid data segments are obtained by filtering from the operational database based on preset filtering conditions. Based on the GA-LM joint optimization algorithm and the effective data segment, the identification parameters of the second-order RC battery model are optimized and identified, and the identification parameter set is output. The vehicle-side battery management system is optimized based on the identification parameters in the identification parameter group, and a predicted battery state is output.

2. The method for online identification and updating of battery model parameters based on cloud collaboration as described in claim 1, characterized in that, The preset screening conditions include current excitation sufficiency conditions, temperature stability conditions, SOC variation range screening conditions, and data continuity conditions. Specifically, the current excitation sufficiency condition requires that the standard deviation of the current within the effective data segment is greater than a preset multiple threshold of the battery capacity; the temperature stability condition requires that the temperature fluctuation value within the effective data segment is less than or equal to a preset temperature threshold; the SOC variation range screening condition selects a first slope interval for lithium iron phosphate batteries and a second slope interval for ternary lithium batteries, with the first slope interval partially overlapping with the second slope interval; and the data continuity condition requires that the time period length corresponding to the effective data segment is greater than or equal to a preset time length.

3. The online identification and updating method for battery model parameters based on cloud collaboration as described in claim 1, characterized in that, The optimization and identification of the identification parameters of the second-order RC battery model based on the GA-LM joint optimization algorithm and the effective data segment, and the output of the identification parameter set, specifically includes: The original parameters of the GA-LM joint optimization algorithm are initialized, including gene encoding method, population size, number of iterations, and individual fitness. The fitness function of the GA-LM joint optimization algorithm is determined based on the individual fitness and the number of individuals in the population, and the fitness function is used to evaluate each individual in the population, and the individuals in the population are sorted according to their fitness. A crossover operation is performed on random individuals in the population to generate offspring individuals, and a mutation operation is performed on each offspring individual to generate new genomes and update the population. If the number of iterations of the GA-LM joint optimization algorithm reaches the preset number of iterations or the fitness of the current individual is greater than the preset fitness, then the iteration process of the GA-LM joint optimization algorithm ends, the individual with the highest fitness in the last generation population is selected as the optimal solution, and the optimization parameters are obtained by decoding to optimize and identify the identification parameters of the second-order RC battery model and output the identification parameter set.

4. The online identification and updating method for battery model parameters based on cloud collaboration as described in claim 3, characterized in that, The method further includes: The identification parameter set, the average temperature of the data segment corresponding to the identification parameter set, and the average state of charge corresponding to the identification parameter set are used to construct an identification parameter triplet. The identification parameter triplet is then written into the parameter mapping relationship diagram corresponding to the battery of the vehicle under test. The parameter mapping relationship diagram is a multi-dimensional lookup table, with the input dimensions being temperature and state of charge, and the output dimension being battery model parameters. The parameter mapping relationship diagram provides the vehicle-side battery management system with the most matching parameter values ​​to achieve online adaptive updating of the model.

5. The online identification and updating method for battery model parameters based on cloud collaboration as described in claim 1, characterized in that, The step of optimizing the vehicle-side battery management system based on the identification parameters in the identification parameter group and outputting a predicted battery state specifically includes: If the parameter mapping relationship diagram of the vehicle under test meets the update conditions, the cloud server sends the identification parameter group to the vehicle-side battery management system. The update conditions include that the difference between the identification parameter group and the original identification parameter group is greater than a preset identification difference value, or that the time of sending by the cloud server is greater than the time of the last sending than a preset update period.

6. The method for online identification and updating of battery model parameters based on cloud collaboration as described in claim 1, characterized in that, The expression corresponding to the second-order RC battery model is: in, U RC1 ( k ) indicates the first k Electrochemical polarization voltage at time, τ 1 represents the time constant corresponding to the electrochemical polarization circuit. R 1 represents the electrochemical polarization resistance. Indicates the sampling time interval. I ( k ) indicates the first k Battery current at any given moment U RC2 ( k ) indicates the first k Concentration polarization voltage at time, τ 2 represents the time constant corresponding to the concentration polarization loop. R 2 represents the concentration polarization internal resistance. U L ( k ) indicates the first k Battery terminal voltage at time, U oc ( k ) No. k The battery open-circuit voltage at any given time. R 0 represents the ohmic internal resistance of the battery.

7. The online identification and updating method for battery model parameters based on cloud collaboration as described in claim 1, characterized in that, The identification parameters include ohmic internal resistance, electrochemical polarization internal resistance, electrochemical polarization capacitance, concentration polarization internal resistance, and concentration polarization capacitance.

8. A cloud-based collaborative online battery model parameter identification and update system, characterized in that, The online identification and update system (1) for battery model parameters includes a data acquisition module (11), a data processing module (12), and an identification and update module (13), wherein, The data acquisition module (11) is used to acquire the battery operation data of the vehicle under test and calculate the initial state of charge of the vehicle under test based on the battery operation data. The data processing module (12) is used to build an operating database on the cloud server based on all battery operating data and the initial state of charge, and to filter from the operating database based on preset filtering conditions to obtain effective data segments; based on the GA-LM joint optimization algorithm and the effective data segments, to optimize and identify the identification parameters of the second-order RC battery model and output the identification parameter group. The identification update module (13) is used to optimize the parameters of the vehicle-side battery management system according to the identification parameters in the identification parameter group and output the predicted battery status.

9. An electronic device, characterized in that, The device includes a processor (21), a memory (25), a user interface (23), and a network interface (24). The memory (25) is used to store instructions. The user interface (23) and the network interface (24) are used to communicate with other devices. The processor (21) is used to execute the instructions stored in the memory (25) to cause the electronic device (2) to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-7.

Images