Battery life optimization method and device, computer equipment, readable storage medium and program product

Abstract

The invention relates to a battery life optimization method and device, computer equipment, a computer readable storage medium and a computer program product. The method includes: acquiring vibration characteristics of a vehicle; establishing a battery health state change rate model, a future health state model, a battery life objective function and a battery operation constraint of the battery based on the vibration characteristics and the health state; constructing a prediction time period, and predicting a future health state change track of the battery in the prediction time period through a prediction controller based on the vibration characteristics and a future health state model; and the trajectory is substituted into a battery life objective function, battery operation constraints are combined, the current optimal power and the current optimal temperature in a prediction time period are solved through a prediction controller, and the operation state of the vehicle is controlled through a vehicle controller based on the current optimal temperature second power. By adopting the method, the relationship between the vehicle vibration and the battery SOH can be accurately established so as to prolong the service life of the battery.

Description

Technical Field

[0001] This application relates to the field of power battery technology, and in particular to a battery life optimization method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Technology

[0002] With the widespread adoption of new energy vehicles, the full lifecycle management of power batteries has become one of the core issues restricting the industry's development. Traditional power battery life prediction models and management strategies mainly focus on factors related to electrochemical characteristics such as temperature, charging and discharging current, and terminal voltage, but generally ignore the mechanical vibrations caused by differences in structural design of different vehicle models (such as battery installation layout and suspension stiffness) and complex road conditions (such as bumpy roads and unpaved roads) in actual use scenarios. Such vibrations have a significant disturbance effect on the integrity of the battery's internal structure and the stability of electrode reactions, thereby accelerating the degradation process of the battery's state of health (SOH).

[0003] Existing research is mostly limited to standardized vibration simulation tests in laboratory environments. The test conditions deviate significantly from the dynamic vibration characteristics in actual road driving, making it difficult to accurately establish the nonlinear mapping relationship between vibration characteristics and battery SOH decay under actual operating conditions for different vehicle models, and thus difficult to effectively improve battery life. Summary of the Invention

[0004] Therefore, it is necessary to provide a battery life optimization method, apparatus, computer equipment, computer-readable storage medium, and computer program product that can accurately establish the mapping relationship between vehicle vibration characteristics and battery SOH decay to improve battery life, in order to address the above-mentioned technical problems.

[0005] In a first aspect, this application provides a method for optimizing battery life, including:

[0006] Acquire initial acceleration data, vehicle structural parameters, battery status parameters, and health status of the vehicle while it is in motion; battery status parameters include the vehicle battery's state of charge, power, and temperature.

[0007] The initial acceleration data is preprocessed to obtain acceleration data; the time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature and state of charge are fused to obtain vibration features;

[0008] Based on vibration characteristics and health status, a future health status model of the battery is established through a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate model, a battery life objective function and battery operation constraints are established; the battery life objective function minimizes the battery health status change rate by optimizing the battery's power and temperature.

[0009] A prediction time period is constructed. Based on vibration characteristics and a future health state model, the prediction controller predicts the trajectory of the battery's future health state during the prediction time period. The trajectory of the future health state is substituted into the battery life objective function. With the goal of minimizing the cumulative change in future battery health and combined with battery operation constraints, the optimal power and temperature control sequence within the prediction time period is obtained by solving the prediction controller.

[0010] The first power and temperature combination in the optimal power and temperature control sequence is taken as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, the vehicle's operating status is controlled by the vehicle controller.

[0011] In one embodiment, the extraction of time-domain and frequency-domain features of the acceleration data includes:

[0012] The effective value, peak value, average value, variance, kurtosis, and skewness of the acceleration data are calculated to obtain the time-domain characteristics; the dominant frequency, total spectral energy, and spectral entropy of the acceleration data are calculated to obtain the frequency-domain characteristics.

[0013] In one embodiment, based on vibration characteristics and health status, a future health status model of the battery is established using a preset algorithm, including:

[0014] Based on vibration characteristics, a battery health state change rate model is established using a preset algorithm; based on the health state and the battery health state change rate model, a future health state model of the battery is established.

[0015] In one embodiment, the method further includes:

[0016] Acquire data samples of the target vehicle; the data samples include the target battery health status and target vibration characteristics; based on the target vibration characteristics, determine the predicted battery health status of the target vehicle through a predictive controller using a future health status model and a battery life objective function; calculate the residual between the target battery health status and the predicted battery health status; if the residual is greater than a preset value, add the data sample to the supplementary sample pool.

[0017] In one embodiment, the method further includes:

[0018] Every preset number of days, or whenever the number of data samples in the supplementary sample pool reaches a preset number, the future health status model is retrained based on the data samples in the supplementary sample pool.

[0019] In one embodiment, establishing battery operating constraints includes:

[0020] Set the power range, temperature range, battery charging rate range, and remaining battery capacity range as battery operation constraints.

[0021] Secondly, this application also provides a battery life optimization device, comprising:

[0022] The acquisition module is used to acquire the vehicle's initial acceleration data, vehicle structural parameters, battery status parameters, and health status during driving; the battery status parameters include the vehicle battery's state of charge, power, and temperature.

[0023] The extraction module is used to preprocess the initial acceleration data to obtain acceleration data; extract the time-domain and frequency-domain features of the acceleration data; and fuse the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature, and state of charge to obtain vibration features.

[0024] A module is established to build a future health state model of the battery based on vibration characteristics and health status using a preset algorithm. The future health state model includes a battery health state change rate model. Based on the battery health state change rate model, a battery life objective function and battery operation constraints are established. The battery life objective function minimizes the battery health state change rate by optimizing the battery's power and temperature.

[0025] The prediction module is used to construct a prediction time period. Based on vibration characteristics and a future health state model, it predicts the trajectory of the battery's future health state change during the prediction time period through a prediction controller. Substituting the trajectory of the future health state change into the battery life objective function and combining it with battery operation constraints, the prediction controller solves for the optimal power and temperature control sequence within the prediction time period.

[0026] The control module is used to take the first power and temperature combination in the optimal power and temperature control sequence as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, the vehicle controller controls the vehicle's operating status.

[0027] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0028] Acquire initial acceleration data, vehicle structural parameters, battery status parameters, and health status of the vehicle while it is in motion; battery status parameters include the vehicle battery's state of charge, power, and temperature.

[0029] The initial acceleration data is preprocessed to obtain acceleration data; the time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature and state of charge are fused to obtain vibration features;

[0030] Based on vibration characteristics and health status, a future health status model of the battery is established through a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate model, a battery life objective function and battery operation constraints are established; the battery life objective function minimizes the battery health status change rate by optimizing the battery's power and temperature.

[0031] A prediction time period is constructed. Based on vibration characteristics and a future health state model, the prediction controller predicts the trajectory of the battery's future health state during the prediction time period. The future health state trajectory is substituted into the battery life objective function. Combined with battery operation constraints, the optimal power and temperature control sequence within the prediction time period is obtained by solving the prediction controller.

[0032] The first power and temperature combination in the optimal power and temperature control sequence is taken as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, the vehicle's operating status is controlled by the vehicle controller.

[0033] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0034] Acquire initial acceleration data, vehicle structural parameters, battery status parameters, and health status of the vehicle while it is in motion; battery status parameters include the vehicle battery's state of charge, power, and temperature.

[0035] The initial acceleration data is preprocessed to obtain acceleration data; the time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature and state of charge are fused to obtain vibration features;

[0036] Based on vibration characteristics and health status, a future health status model of the battery is established through a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate model, a battery life objective function and battery operation constraints are established; the battery life objective function minimizes the battery health status change rate by optimizing the battery's power and temperature.

[0037] A prediction time period is constructed. Based on vibration characteristics and a future health state model, the prediction controller predicts the trajectory of the battery's future health state during the prediction time period. The future health state trajectory is substituted into the battery life objective function. Combined with battery operation constraints, the optimal power and temperature control sequence within the prediction time period is obtained by solving the prediction controller.

[0038] The first power and temperature combination in the optimal power and temperature control sequence is taken as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, the vehicle's operating status is controlled by the vehicle controller.

[0039] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0040] Acquire initial acceleration data, vehicle structural parameters, battery status parameters, and health status of the vehicle while it is in motion; battery status parameters include the vehicle battery's state of charge, power, and temperature.

[0041] The initial acceleration data is preprocessed to obtain acceleration data; the time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature and state of charge are fused to obtain vibration features;

[0042] Based on vibration characteristics and health status, a future health status model of the battery is established through a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate model, a battery life objective function and battery operation constraints are established; the battery life objective function minimizes the battery health status change rate by optimizing the battery's power and temperature.

[0043] A prediction time period is constructed. Based on vibration characteristics and a future health state model, the prediction controller predicts the trajectory of the battery's future health state during the prediction time period. The future health state trajectory is substituted into the battery life objective function. Combined with battery operation constraints, the optimal power and temperature control sequence within the prediction time period is obtained by solving the prediction controller.

[0044] The first power and temperature combination in the optimal power and temperature control sequence is taken as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, the vehicle's operating status is controlled by the vehicle controller.

[0045] The aforementioned battery life optimization method, apparatus, computer equipment, computer-readable storage medium, and computer program product first acquire initial acceleration data, vehicle structural parameters, battery state parameters, and health status of the vehicle during operation; the battery state parameters include the vehicle battery's state of charge, power, and temperature; the initial acceleration data is preprocessed to obtain acceleration data; time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature, and state of charge are fused to obtain vibration features; then, based on the vibration features and health status, a future health status model of the battery is established using a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate... The model establishes a battery life objective function and battery operation constraints. The battery life objective function minimizes the rate of change of battery health status by optimizing battery power and temperature. Then, a prediction time period is constructed. Based on vibration characteristics and the future health status model, a predictive controller predicts the trajectory of future battery health status changes within the prediction time period. Substituting the future health status change trajectory into the battery life objective function, combined with battery operation constraints, the predictive controller solves for the optimal power and temperature control sequence within the prediction time period. Finally, the first power and temperature combination in the optimal power and temperature control sequence is used as the current optimal power and temperature. Based on the current optimal power and temperature, the vehicle controller controls the vehicle's operating state. This application integrates vehicle structure, vibration characteristics, and battery operation data to construct a comprehensive feature model. Based on a preset algorithm, a mapping model between vibration and state of health (SOH) decay is established. Combined with a predictive controller, real-time adaptive optimization control is achieved, which can accurately establish the mapping relationship between vehicle vibration characteristics and battery SOH decay, thereby controlling the vehicle's state to improve battery life. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart illustrating a battery life optimization method in one embodiment;

[0048] Figure 2 This is a detailed flowchart illustrating a battery life optimization method in one embodiment;

[0049] Figure 3 This is a structural block diagram of a battery life optimization device in one embodiment;

[0050] Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0052] In one embodiment, such as Figure 1 As shown, a battery life optimization method is provided. This embodiment illustrates the method's application to a terminal. It is understood that this method can also be applied to a server, and further to a system including both a terminal and a server, and implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0053] Step 102: Obtain the initial acceleration data of the vehicle while it is in motion, vehicle structural parameters, battery status parameters and health status; the battery status parameters include the state of charge, power and temperature of the vehicle battery.

[0054] The initial acceleration data is collected in real time by a triaxial accelerometer mounted on the vehicle, with a sampling frequency of 50-200Hz, covering vibration acceleration signals in three dimensions during vehicle movement: longitudinal (front-back, x-axis), lateral (left-right, y-axis), and vertical (up-down, z-axis). It comprehensively captures vibration characteristics caused by different road conditions (such as paved roads, unpaved roads, speed bumps, etc.) and driving behaviors (such as rapid acceleration and sudden braking). Vehicle structural parameters include battery pack installation layout (such as chassis embedded or trunk placement), suspension stiffness level, vehicle weight distribution, battery pack fixing method, and buffer structure parameters, which are used to quantify the impact of vehicle structure differences on vibration transmission.

[0055] Optionally, preprocessing includes filtering, static offset removal, and normalization. Filtering uses a low-pass filter (e.g., a Butterworth filter) for noise suppression, setting the cutoff frequency to 50Hz to retain key vibration frequencies below 50Hz related to battery aging, while filtering out high-frequency interference signals such as sensor electronic noise and impacts from minor road imperfections. Static offset removal is performed, using a baseline calibration algorithm to eliminate fixed offsets caused by gravity and mechanical mounting biases, ensuring the data focuses on dynamic vibration changes during vehicle operation. Normalization is implemented, using a max-min normalization method to map the processed acceleration data to a unified amplitude range of [0,1], eliminating the influence of dimensional differences and amplitude spans in vibration signals from different dimensions, providing a standardized data foundation for subsequent time-frequency feature extraction and multi-dimensional data fusion calculations.

[0056] Step 104: Preprocess the initial acceleration data to obtain acceleration data; extract the time domain features and frequency domain features of the acceleration data; fuse the vehicle structural parameters, time domain features, frequency domain features, power, temperature and state of charge to obtain vibration features.

[0057] Optionally, the vehicle structural parameters (including vehicle mass, suspension stiffness, damping coefficient, battery module center of gravity position, battery installation height, and vehicle structure coding), time-domain features, frequency-domain features, power, battery temperature, and SOC (State of Charge, remaining battery capacity) are first concatenated into a one-dimensional feature vector in a preset order. Then, the continuous features in this vector (such as power, temperature, and time / frequency domain feature values) are Z-score standardized, and the discrete vehicle structural parameters are transformed using one-hot encoding to eliminate the dimensional differences and numerical spans of different types of features. Finally, a high-dimensional vibration feature vector with unified dimensions and standardized values ​​is formed, which not only retains the core information of various original features but also achieves effective coupling between features, providing a comprehensive and accurate input basis for the subsequent construction of the battery's future health status model.

[0058] Step 106: Based on vibration characteristics and health status, establish a future health status model of the battery using a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate model, establish a battery life objective function and battery operation constraints; the battery life objective function minimizes the battery health status change rate by optimizing the battery's power and temperature.

[0059] Optionally, the default algorithm uses the XGBoost (Extreme Gradient Boosting Regression Algorithm). Building a model of the battery's future health status involves training the model.

[0060] Specifically, a nonlinear regression model is established to fit the relationship between SOH and vibration state:

[0061]

[0062] in It is a vibration characteristic. This represents the random perturbation term. Considering the strong nonlinear relationship, a supervised machine learning method is used to model this function. Applicable regression models include... XGBoost uses an ensemble method with weighted regression trees:

[0063] .

[0064] Each regression tree It maps vibrational characteristics to the rate of change of SOH and has a strong generalization ability.

[0065] Step 108: Construct a prediction time period. Based on vibration characteristics and a future health state model, predict the trajectory of the battery's future health state during the prediction time period using a predictive controller. Substitute the trajectory of the future health state into the battery life objective function, and combine it with battery operation constraints to obtain the optimal power and temperature control sequence within the prediction time period through the predictive controller.

[0066] The prediction time period is a fixed-length discretized time interval [t, t+H], where t represents the current control time and H is the prediction time domain length (typically 10-60 seconds, depending on the real-time requirements of vehicle control). The time step within the interval is divided according to the control cycle. (Typically 0.1-0.5 seconds), forming H / A series of continuous prediction nodes ensures that the trend changes in battery state of health (SOH) can be captured, while avoiding the accumulation of errors caused by excessively long prediction times.

[0067] Specifically, the prediction of future health status trajectories is based on short-term trend prediction of vibration characteristics and centered on a future health status model. The specific process is as follows: Vibration characteristic prediction: Based on the historical vibration characteristic vectors of the current period and the past 3-5 control cycles, the vibration characteristics of each node within the prediction period are predicted using the moving average method or a lightweight LSTM model. (τ=0, 1, ..., H / ), ensuring that the features are consistent with the actual road condition trend; initial state substitution: using the current battery health state SOH(t) as the prediction starting point, inputting it into the future health state model; point-by-point iterative calculation: based on the state transition equation:

[0068]

[0069] Where f() is the health status change rate model, the SOH value of each prediction node is calculated in sequence, and finally a complete decay trajectory from SOH(t) to SOH(t+H) is formed, which intuitively reflects the battery aging trend under different control strategies.

[0070] Furthermore, the objective function is:

[0071]

[0072] The optimal power and temperature control sequence is solved by a predictive controller (i.e., a model predictive controller, MPC) optimizing the objective function within the constraint framework. The steps are as follows: Define the optimization variables: define the control quantity of each node within the prediction time period ( ), of which battery charging and discharging power Operating temperature Let these be the optimization variables; transform the battery operation constraints into mathematical inequalities, and simultaneously link power and SOC state through the SOC prediction model to ensure the constraints are feasible; solve the objective function: substitute the predicted trajectory into the objective function to obtain:

[0073]

[0074] By calling an online optimizer (Nonlinear Programming (NLP) or Quadratic Programming (QP) to adapt to the nonlinear characteristics of the model), and under the premise of satisfying all constraints, the combination of control variables that minimizes J is solved, thus obtaining the optimal power and temperature control sequence.

[0075] Step 110: Take the first power and temperature combination in the optimal power and temperature control sequence as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, control the vehicle's operating status through the vehicle controller.

[0076] The optimal power and temperature control sequence contains the theoretically optimal control parameters for each time step within the prediction period. However, considering the dynamic randomness of vehicle driving conditions, vibration characteristics, and battery status, only the first power and temperature combination in the sequence is selected as the current optimal power and current optimal temperature. This combination is the optimal solution obtained based on the actual state data at the current moment, and has the highest adaptability and reliability. The control parameters for subsequent time steps are only used as prediction references and are not directly executed.

[0077] Optionally, vehicle operating state control based on the current optimal power and temperature is implemented through the vehicle controller (including collaborative control between the vehicle control unit (VCU) and the battery management system (BMS)). The specific process is as follows: An example of a control strategy mapping model is given below. The maximum output power is limited... Mapped to maximum allowable current :

[0078]

[0079] fan / pump( Control function (based on temperature target):

[0080]

[0081]

[0082] Establish a strategy deployment mechanism. The controller runs optimization and MPC modules locally, and the generated control strategies are encapsulated in CAN protocol and sent to the BMS and thermal management system via the CAN (Controller Area Network) bus. The VCU / BMS updates operating parameters according to the instructions. Given this background, safety redundancy design is required. The system is equipped with an anomaly detection mechanism; if vibration data or optimization fails, or the model output is unreasonable, it will revert to the default strategy and trigger an alarm.

[0083] Furthermore, the system of this invention is composed of a software system and a hardware system working together to form a closed-loop intelligent management platform for new energy vehicles, encompassing "vibration sensing—battery life prediction—control optimization—feedback execution." It boasts high performance, high adaptability, and engineering feasibility, achieving intelligent health management throughout the entire lifecycle of the power battery through deep integration of software and hardware. The software system adopts a modular, layered design, comprising five core functional modules, each with its own function and working in synergy: the vibration data processing module first performs filtering and windowing processing on the raw triaxial acceleration data, accurately extracting multi-dimensional time-domain and frequency-domain features such as effective values, peak values, and spectral entropy, and standardizing them into a standardized model input vector, providing high-quality data support for subsequent predictions; the SOH prediction model running module loads the cloud-trained and optimized XGBoost regression model, estimating the attenuation effect of vibration conditions on the health status of the power battery online based on real-time vibration characteristics and the current state of the battery, outputting an accurate SOH change rate; the MPC optimization control module minimizes the prediction interval. With the internal state of harm (SOH) degradation rate as the core objective, and combined with battery operating constraints, the optimal combination of control variables is solved online, including key parameters such as charging rate, thermal management level, and power output limit. The strategy mapping and command generation module transforms the optimal control parameters obtained from the MPC solution into standardized control commands that conform to the vehicle communication protocol, ensuring that the commands can be directly recognized by the underlying controller. The data synchronization and anomaly monitoring module realizes real-time data interaction with the cloud platform, covering functions such as uploading operating data, updating model parameters, and version management. It also has a built-in fault diagnosis mechanism that triggers alarms for abnormal data acquisition and model inference failure, ensuring long-term stable operation of the system.The hardware system employs automotive-grade components and a high-reliability design, divided into three core units to meet the stringent requirements of automotive scenarios: The vibration acquisition unit uses a high-precision triaxial MEMS (Micro-Electro-Mechanical Systems) accelerometer with a range covering ±2g to ±16g. It is rigidly mounted on the battery pack housing or battery structural connection points to ensure accurate capture of vibration transmission characteristics. The sensor also features temperature compensation and electromagnetic interference resistance, with a recommended sampling frequency of at least 500Hz, capable of capturing both high and low frequency vibration characteristics and adapting to complex road conditions. The data processing and control unit is equipped with an automotive-grade main control chip, possessing efficient floating-point arithmetic capabilities and low power consumption. It can locally complete key tasks such as feature extraction, model inference, and MPC optimization, avoiding data transmission delays. It also has a built-in Flash storage module for caching operational data and fault information, supports OTA remote upgrades, and features a model encryption anti-reverse engineering mechanism to ensure technical security. The communication and deployment module uses CAN-FD (Controller Area Network with Flexible...) It establishes communication connections with the vehicle's BMS and VCU through the Data-Rate (Flexible Data Rate Controller Area Network) or CAN2.0 (Version 2.0 Controller Area Network) protocol. It achieves high and low voltage circuit isolation through isolation chips, and suppresses signal interference with common mode inductors and terminating resistors to ensure highly reliable transmission and writing of control commands, thereby realizing real-time dynamic regulation of the power battery.

[0084] The aforementioned battery life optimization method, apparatus, computer equipment, computer-readable storage medium, and computer program product first acquire initial acceleration data, vehicle structural parameters, battery state parameters, and health status of the vehicle during operation; the battery state parameters include the vehicle battery's state of charge, power, and temperature; the initial acceleration data is preprocessed to obtain acceleration data; time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature, and state of charge are fused to obtain vibration features; then, based on the vibration features and health status, a future health status model of the battery is established using a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate... The model establishes a battery life objective function and battery operation constraints. The battery life objective function minimizes the rate of change of battery health status by optimizing battery power and temperature. Then, a prediction time period is constructed. Based on vibration characteristics and the future health status model, a predictive controller predicts the trajectory of future battery health status changes within the prediction time period. Substituting the future health status change trajectory into the battery life objective function, combined with battery operation constraints, the predictive controller solves for the optimal power and temperature control sequence within the prediction time period. Finally, the first power and temperature combination in the optimal power and temperature control sequence is used as the current optimal power and temperature. Based on the current optimal power and temperature, the vehicle controller controls the vehicle's operating state. This application integrates vehicle structure, vibration characteristics, and battery operation data to construct a comprehensive feature model. Based on a preset algorithm, a mapping model between vibration and state of health (SOH) decay is established. Combined with a predictive controller, real-time adaptive optimization control is achieved, which can accurately establish the mapping relationship between vehicle vibration characteristics and battery SOH decay, thereby controlling the vehicle's state to improve battery life.

[0085] In an exemplary embodiment, extracting the time-domain and frequency-domain features of the acceleration data includes:

[0086] The effective value, peak value, average value, variance, kurtosis, and skewness of the acceleration data are calculated to obtain the time-domain characteristics; the dominant frequency, total spectral energy, and spectral entropy of the acceleration data are calculated to obtain the frequency-domain characteristics.

[0087] For example, the acceleration signal in each direction (i.e. Representative statistical features are extracted separately. These features are divided into time-domain features and frequency-domain features. The process of extracting time-domain features is as follows: i represents x, y, and z.

[0088] Effective value (RMS): Represents the energy level of the signal, calculated using the following formula:

[0089]

[0090] Peak value: Represents the maximum amplitude of the signal, i.e.:

[0091]

[0092] Mean: Reflects the center of vibration offset.

[0093]

[0094] Variance: Measures the degree of fluctuation.

[0095]

[0096] Kurtosis: Describes the spikes (or bruises) in a signal.

[0097]

[0098] Skewness: describes the degree of left-right asymmetry in the distribution of data.

[0099]

[0100] The process of extracting time-domain features is as follows: Perform a Fast Fourier Transform (FFT) on the acceleration signal in each direction:

[0101] =FFT

[0102] Key characteristics of frequency domain computation:

[0103] Clock speed:

[0104]

[0105] Total Spectral Energy:

[0106]

[0107] Spectral entropy describes the "degree of disorder" in frequency distribution; a higher entropy indicates more complex vibrations.

[0108]

[0109] Among them, probability distribution as follows:

[0110]

[0111] All of the above features are used as input variables in subsequent modeling.

[0112] In this embodiment, by extracting time-domain and frequency-domain features from the three-axis acceleration data during vehicle operation, the essential vibration characteristics of the vehicle under actual working conditions can be comprehensively characterized from multiple dimensions such as vibration energy intensity, impact characteristics, stability, and frequency distribution. Through independent extraction and feature integration of three-axis data, the system can capture vibration information in all directions (longitudinal, lateral, and vertical) without omission, laying a solid foundation for subsequent feature fusion with vehicle structural parameters and battery operating status (power, temperature, SOH).

[0113] In one embodiment, based on vibration characteristics and health status, a future health status model of the battery is established using a preset algorithm, including:

[0114] Based on vibration characteristics, a battery health state change rate model is established using a preset algorithm; based on the health state and the battery health state change rate model, a future health state model of the battery is established.

[0115] For example, a nonlinear regression model is established to fit the relationship between SOH and vibration state:

[0116]

[0117] in Represents random perturbation terms; XGBoost (the default algorithm) uses a weighted regression tree ensemble method:

[0118]

[0119] Among them, each regression tree Mapping vibration characteristics to the rate of change of SOH. The output is the rate of change of SOH within a certain time window, yielding the battery health state rate model:

[0120]

[0121] A model of future health status can be obtained by continuous integration:

[0122] .

[0123] In this embodiment, a step-by-step construction strategy combined with the XGBoost regression algorithm is used to accurately establish the mapping relationship between vibration characteristics and battery SOH decay rate, realize continuous prediction of future SOH decay trajectory, adapt to dynamic changes in operating conditions, provide an accurate prediction basis for online optimization control, and specifically offset the problem of accelerated battery aging caused by vibration.

[0124] In one embodiment, the method further includes:

[0125] Acquire data samples of the target vehicle; the data samples include the target battery health status and target vibration characteristics; based on the target vibration characteristics, determine the predicted battery health status of the target vehicle through a predictive controller using a future health status model and a battery life objective function; calculate the residual between the target battery health status and the predicted battery health status; if the residual is greater than a preset value, add the data sample to the supplementary sample pool.

[0126] For example, during vehicle operation, the system continuously collects vibration data, battery operation data, and structural parameters of the target vehicle. After preprocessing, feature extraction, and fusion, target vibration features are formed. The target battery health status (SOH) for the corresponding time period is calculated by the battery management system (BMS), thereby constructing a data sample containing the target vibration features and the target battery health status. Subsequently, the target vibration features are input into the pre-trained future health status model. Combined with the battery life objective function, the predicted value of the battery health status for that time period is obtained by the predictive controller. The absolute residual is calculated based on the target battery health status and the predicted value. If the residual is greater than a preset threshold (e.g., 0.01% / cycle, set according to battery aging characteristics and model accuracy requirements), it indicates that the current model does not fit the mapping relationship between vibration and SOH decay under this operating condition well enough. The system automatically adds the data sample to the supplementary sample pool.

[0127] In this embodiment, by triggering a data acquisition mechanism, the actual battery health status and corresponding vibration characteristic data samples of the target vehicle are periodically acquired. The health status is predicted using the constructed future health status model and predictive controller. Samples with excessive model fitting deviation are screened out by residual comparison and included in the supplementary sample pool. This ensures the timeliness and operating condition coverage of the supplementary data, and can also specifically compensate for the model's insufficient fitting in long-term driving or specific mileage stages. It provides accurate targeted data for subsequent incremental training of the model, continuously improves the model's adaptability to complex vibration conditions throughout the entire life cycle and the accuracy of SOH prediction, and ensures the long-term effectiveness and reliability of the battery life optimization control strategy.

[0128] In one embodiment, the method further includes:

[0129] Every preset number of days, or whenever the number of data samples in the supplementary sample pool reaches a preset number, the future health status model is retrained based on the data samples in the supplementary sample pool.

[0130] To ensure the model maintains high prediction accuracy and generalization ability across different vehicle models and operating conditions, the system employs a periodic retraining mechanism. Specifically, every 30 days or after accumulating 100,000 kilometers of vehicle driving data, the backend platform automatically aggregates triaxial vibration data and SOH label data uploaded from each vehicle and uses a batch training framework to retrain and optimize the main model. After training is complete, the system generates an encrypted model file and pushes it to the terminal controllers of each vehicle, enabling online updates and replacements of the model.

[0131] For example, every 30 days or after accumulating 100,000 kilometers of vehicle driving data, or after accumulating 500 data points in the sample pool, the system initiates an incremental training process for the future health status model: First, the data samples in the supplementary sample pool are subjected to consistency verification and standardization preprocessing, outliers and missing values ​​are removed, and feature and label data are calibrated; then, based on the incremental learning characteristics of the XGBoost regression algorithm, the preprocessed supplementary samples are used as the incremental training set based on the original trained model, and only the model parameters are updated; after training is completed, the supplementary sample pool is cleared to accumulate a new round of targeted data, continuously optimizing the model's adaptability to complex vibration conditions and the accuracy of SOH prediction.

[0132] In this embodiment, the future health status model is incrementally trained periodically based on targeted data in the supplementary sample pool. This not only allows for efficient integration of new operating condition information without full retraining, but also addresses the model's insufficient fitting in complex vibration scenarios, continuously optimizing the model's prediction accuracy and operating condition adaptability. This ensures that the battery's future health status prediction and lifespan optimization control strategy maintains high reliability and effectiveness throughout the vehicle's entire life cycle.

[0133] In one embodiment, establishing battery operating constraints includes:

[0134] Set the power range, temperature range, battery charging rate range, and remaining battery capacity range as battery operation constraints.

[0135] For example, the battery operating constraints are:

[0136]

[0137] in, It's the charging rate. This is the remaining battery power.

[0138] In this embodiment, a multi-dimensional battery operation constraint system is constructed, which includes power constraints, temperature constraints, charging rate constraints, and remaining charge (SOC) constraints. This system clearly defines the boundary conditions for the safe and stable operation of the battery. It provides a rigid boundary for optimizing the battery life objective function and ensures the safety of battery operation from four core dimensions: power, temperature, charging rate, and remaining charge. This ensures that the optimized power and temperature control sequence extends battery life without exceeding the battery's own operating limits, achieving a synergistic unity between life extension and safe operation.

[0139] In one exemplary embodiment, such as Figure 2 As shown, a battery life optimization method includes:

[0140] Acquire initial acceleration data, vehicle structural parameters, battery state parameters, and health status of the vehicle during operation. Battery state parameters include the battery's state of charge, power, and temperature. Preprocess the initial acceleration data to obtain acceleration data. Extract time-domain and frequency-domain features from the acceleration data. Fusion the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature, and health status to obtain vibration characteristics. Establish a nonlinear regression model to fit the relationship between SOH (State of Charge) and vibration state.

[0141]

[0142] in Represents random perturbation terms; XGBoost (the default algorithm) uses a weighted regression tree ensemble method:

[0143]

[0144] Among them, each regression tree Mapping vibration characteristics to the rate of change of SOH. The output is the rate of change of SOH within a certain time window, yielding the battery health state rate model:

[0145]

[0146] A model of future health status can be obtained by continuous integration:

[0147] .

[0148] Based on the battery health state change rate model, a battery life objective function and battery operation constraints are established; the objective function is:

[0149]

[0150] Battery operating constraints are:

[0151]

[0152] Where P is power and T is temperature. It's the charging rate. This refers to the remaining battery power. The battery life objective function minimizes the rate of change in battery health by optimizing battery power and temperature. A prediction time period [t, t+H] is constructed, where t represents the current control moment and H is the prediction time domain length (typically 10-60 seconds, depending on the real-time requirements of vehicle control). The time step within this interval is divided according to the control cycle. (Typically 0.1-0.5 seconds), forming H / A series of continuous prediction nodes. Based on vibration characteristics and a future health state model, the predictive controller predicts the trajectory of the battery's future health state over the prediction period. Specifically, the prediction of the future health state trajectory is based on the short-term trend prediction of vibration characteristics and the future health state model. The specific process is as follows: Vibration characteristic prediction: Based on the historical vibration characteristic vectors of the current period and the past 3-5 control cycles, the moving average method or a lightweight LSTM model is used to predict the vibration characteristics of each node within the prediction period. (τ=0,1,...,H / ), ensuring that the features are consistent with the actual road condition trend; initial state substitution: using the current battery health state SOH(t) as the prediction starting point, inputting it into the future health state model; point-by-point iterative calculation: based on the state transition equation:

[0153]

[0154] Where (f() is the health state change rate model), the SOH value of each prediction node is calculated sequentially, ultimately forming a complete decay trajectory from SOH(t) to SOH(t+H), intuitively reflecting the battery aging trend under different control strategies. Substituting the future health state change trajectory into the battery life objective function, combined with battery operation constraints, the optimal power and temperature control sequence within the prediction time period is obtained through the predictive controller. The optimal power and temperature control sequence is solved by the predictive controller (i.e., model predictive controller, MPC) within the constraint framework, optimizing the objective function. The steps are as follows: Define optimization variables: The control quantity of each node within the prediction time period ( ), of which battery charging and discharging power Operating temperature Let these be the optimization variables; transform the battery operation constraints into mathematical inequalities, and simultaneously link power and SOC state through the SOC prediction model to ensure the constraints are feasible; solve the objective function: substitute the predicted trajectory into the objective function to obtain:

[0155] .

[0156] An online optimizer (Nonlinear Programming (NLP) or Quadratic Programming (QP) to adapt to the nonlinear characteristics of the model) is invoked to solve for the combination of control variables that minimizes J, thus obtaining the optimal power and temperature control sequence, while satisfying all constraints. The first power and temperature combination in the optimal power and temperature control sequence is used as the current optimal power and temperature. Based on the current optimal power and temperature, the vehicle controller controls the vehicle's operating state. The method also includes: acquiring data samples of the target vehicle; the data samples include the target battery health state and target vibration characteristics; based on the target vibration characteristics, the predictive controller determines the battery health state prediction of the target vehicle using a future health state model and a battery life objective function; calculating the residual between the target battery health state and the battery health state prediction; if the residual is greater than a preset value, the data sample is added to a supplementary sample pool. Every preset number of days, or whenever the data samples in the supplementary sample pool reach a preset number, the future health state model is retrained based on the data samples in the supplementary sample pool.

[0157] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0158] In one exemplary embodiment, such as Figure 3 As shown, a battery life optimization device is provided, including: an acquisition module 301, an extraction module 302, an establishment module 303, a prediction module 304, and a control module 305, wherein:

[0159] The acquisition module is used to acquire the vehicle's initial acceleration data, vehicle structural parameters, battery status parameters, and health status during driving; the battery status parameters include the vehicle battery's state of charge, power, and temperature.

[0160] The extraction module is used to preprocess the initial acceleration data to obtain acceleration data; extract the time-domain and frequency-domain features of the acceleration data; and fuse the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature, and state of charge to obtain vibration features.

[0161] A module is established to build a future health state model of the battery based on vibration characteristics and health status using a preset algorithm. The future health state model includes a battery health state change rate model. Based on the battery health state change rate model, a battery life objective function and battery operation constraints are established. The battery life objective function minimizes the battery health state change rate by optimizing the battery's power and temperature.

[0162] The prediction module is used to construct a prediction time period. Based on vibration characteristics and a future health state model, it predicts the trajectory of the battery's future health state change during the prediction time period through a prediction controller. Substituting the trajectory of the future health state change into the battery life objective function and combining it with battery operation constraints, the prediction controller solves for the optimal power and temperature control sequence within the prediction time period.

[0163] The control module is used to take the first power and temperature combination in the optimal power and temperature control sequence as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, the vehicle controller controls the vehicle's operating status.

[0164] In one embodiment, the extraction module is further configured to:

[0165] The effective value, peak value, average value, variance, kurtosis, and skewness of the acceleration data are calculated to obtain the time-domain characteristics; the dominant frequency, total spectral energy, and spectral entropy of the acceleration data are calculated to obtain the frequency-domain characteristics.

[0166] In one embodiment, the establishment module is further configured to:

[0167] Based on vibration characteristics, a battery health state change rate model is established using a preset algorithm; based on the health state and the battery health state change rate model, a future health state model of the battery is established.

[0168] In one embodiment, the establishment module is further configured to:

[0169] Acquire data samples of the target vehicle; the data samples include the target battery health status and target vibration characteristics; based on the target vibration characteristics, determine the predicted battery health status of the target vehicle through a predictive controller using a future health status model and a battery life objective function; calculate the residual between the target battery health status and the predicted battery health status; if the residual is greater than a preset value, add the data sample to the supplementary sample pool.

[0170] In one embodiment, the establishment module is further configured to:

[0171] Every preset number of days, or whenever the number of data samples in the supplementary sample pool reaches a preset number, the future health status model is retrained based on the data samples in the supplementary sample pool.

[0172] In one embodiment, the establishment module is further configured to:

[0173] Set the power range, temperature range, battery charging rate range, and remaining battery capacity range as battery operation constraints.

[0174] Each module in the aforementioned battery life optimization device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0175] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores initial acceleration data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a battery life optimization method.

[0176] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0177] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0178] Acquire initial acceleration data, vehicle structural parameters, battery status parameters, and health status of the vehicle while it is in motion; battery status parameters include the vehicle battery's state of charge, power, and temperature.

[0179] The initial acceleration data is preprocessed to obtain acceleration data; the time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature and state of charge are fused to obtain vibration features;

[0180] Based on vibration characteristics and health status, a future health status model of the battery is established through a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate model, a battery life objective function and battery operation constraints are established; the battery life objective function minimizes the battery health status change rate by optimizing the battery's power and temperature.

[0181] A prediction time period is constructed. Based on vibration characteristics and a future health state model, the prediction controller predicts the trajectory of the battery's future health state during the prediction time period. The future health state trajectory is substituted into the battery life objective function. Combined with battery operation constraints, the optimal power and temperature control sequence within the prediction time period is obtained by solving the prediction controller.

[0182] The first power and temperature combination in the optimal power and temperature control sequence is taken as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, the vehicle's operating status is controlled by the vehicle controller.

[0183] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0184] The effective value, peak value, average value, variance, kurtosis, and skewness of the acceleration data are calculated to obtain the time-domain characteristics; the dominant frequency, total spectral energy, and spectral entropy of the acceleration data are calculated to obtain the frequency-domain characteristics.

[0185] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0186] Based on vibration characteristics, a battery health state change rate model is established using a preset algorithm; based on the health state and the battery health state change rate model, a future health state model of the battery is established.

[0187] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0188] Acquire data samples of the target vehicle; the data samples include the target battery health status and target vibration characteristics; based on the target vibration characteristics, determine the predicted battery health status of the target vehicle through a predictive controller using a future health status model and a battery life objective function; calculate the residual between the target battery health status and the predicted battery health status; if the residual is greater than a preset value, add the data sample to the supplementary sample pool.

[0189] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0190] Every preset number of days, or whenever the number of data samples in the supplementary sample pool reaches a preset number, the future health status model is retrained based on the data samples in the supplementary sample pool.

[0191] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0192] Set the power range, temperature range, battery charging rate range, and remaining battery capacity range as battery operation constraints.

[0193] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0194] Acquire initial acceleration data, vehicle structural parameters, battery status parameters, and health status of the vehicle while it is in motion; battery status parameters include the vehicle battery's state of charge, power, and temperature.

[0195] The initial acceleration data is preprocessed to obtain acceleration data; the time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature and state of charge are fused to obtain vibration features;

[0196] Based on vibration characteristics and health status, a future health status model of the battery is established through a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate model, a battery life objective function and battery operation constraints are established; the battery life objective function minimizes the battery health status change rate by optimizing the battery's power and temperature.

[0197] A prediction time period is constructed. Based on vibration characteristics and a future health state model, the prediction controller predicts the trajectory of the battery's future health state during the prediction time period. The future health state trajectory is substituted into the battery life objective function. Combined with battery operation constraints, the optimal power and temperature control sequence within the prediction time period is obtained by solving the prediction controller.

[0198] The first power and temperature combination in the optimal power and temperature control sequence is taken as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, the vehicle's operating status is controlled by the vehicle controller.

[0199] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0200] The effective value, peak value, average value, variance, kurtosis, and skewness of the acceleration data are calculated to obtain the time-domain characteristics; the dominant frequency, total spectral energy, and spectral entropy of the acceleration data are calculated to obtain the frequency-domain characteristics.

[0201] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0202] Based on vibration characteristics, a battery health state change rate model is established using a preset algorithm; based on the health state and the battery health state change rate model, a future health state model of the battery is established.

[0203] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0204] Acquire data samples of the target vehicle; the data samples include the target battery health status and target vibration characteristics; based on the target vibration characteristics, determine the predicted battery health status of the target vehicle through a predictive controller using a future health status model and a battery life objective function; calculate the residual between the target battery health status and the predicted battery health status; if the residual is greater than a preset value, add the data sample to the supplementary sample pool.

[0205] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0206] Every preset number of days, or whenever the number of data samples in the supplementary sample pool reaches a preset number, the future health status model is retrained based on the data samples in the supplementary sample pool.

[0207] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0208] Set the power range, temperature range, battery charging rate range, and remaining battery capacity range as battery operation constraints.

[0209] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0210] Acquire initial acceleration data, vehicle structural parameters, battery status parameters, and health status of the vehicle while it is in motion; battery status parameters include the vehicle battery's state of charge, power, and temperature.

[0211] The initial acceleration data is preprocessed to obtain acceleration data; the time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, time-domain features, frequency-domain features, power, temperature and state of charge are fused to obtain vibration features;

[0212] Based on vibration characteristics and health status, a future health status model of the battery is established through a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate model, a battery life objective function and battery operation constraints are established; the battery life objective function minimizes the battery health status change rate by optimizing the battery's power and temperature.

[0213] A prediction time period is constructed. Based on vibration characteristics and a future health state model, the prediction controller predicts the trajectory of the battery's future health state during the prediction time period. The future health state trajectory is substituted into the battery life objective function. Combined with battery operation constraints, the optimal power and temperature control sequence within the prediction time period is obtained by solving the prediction controller.

[0214] The first power and temperature combination in the optimal power and temperature control sequence is taken as the current optimal power and current optimal temperature; based on the current optimal power and current optimal temperature, the vehicle's operating status is controlled by the vehicle controller.

[0215] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0216] The effective value, peak value, average value, variance, kurtosis, and skewness of the acceleration data are calculated to obtain the time-domain characteristics; the dominant frequency, total spectral energy, and spectral entropy of the acceleration data are calculated to obtain the frequency-domain characteristics.

[0217] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0218] Based on vibration characteristics, a battery health state change rate model is established using a preset algorithm; based on the health state and the battery health state change rate model, a future health state model of the battery is established.

[0219] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0220] Acquire data samples of the target vehicle; the data samples include the target battery health status and target vibration characteristics; based on the target vibration characteristics, determine the predicted battery health status of the target vehicle through a predictive controller using a future health status model and a battery life objective function; calculate the residual between the target battery health status and the predicted battery health status; if the residual is greater than a preset value, add the data sample to the supplementary sample pool.

[0221] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0222] Every preset number of days, or whenever the number of data samples in the supplementary sample pool reaches a preset number, the future health status model is retrained based on the data samples in the supplementary sample pool.

[0223] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0224] Set the power range, temperature range, battery charging rate range, and remaining battery capacity range as battery operation constraints.

[0225] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0226] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0227] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for optimizing battery life, characterized in that, The method includes: Acquire initial acceleration data, vehicle structural parameters, battery status parameters, and health status of the vehicle while it is in motion; the battery status parameters include the vehicle battery's state of charge, power, and temperature. The initial acceleration data is preprocessed to obtain acceleration data; the time-domain and frequency-domain features of the acceleration data are extracted; the vehicle structural parameters, the time-domain features, the frequency-domain features, the power, the temperature, and the state of charge are fused to obtain vibration features; Based on the vibration characteristics and the health status, a future health status model of the battery is established through a preset algorithm; the future health status model includes a battery health status change rate model; based on the battery health status change rate model, a battery life objective function and battery operation constraints are established; the battery life objective function minimizes the battery health status change rate by optimizing the battery's power and temperature. A prediction time period is constructed. Based on the vibration characteristics and the future health state model, the prediction controller predicts the trajectory of the battery's future health state change during the prediction time period. The future health state change trajectory is substituted into the battery life objective function. With the goal of minimizing the cumulative change in future battery health and combined with battery operation constraints, the prediction controller solves for the optimal power and temperature control sequence within the prediction time period. The first power and temperature combination in the optimal power and temperature control sequence is taken as the current optimal power and current optimal temperature; based on the current optimal power and the current optimal temperature, the vehicle's operating state is controlled by the vehicle controller.

2. The method according to claim 1, characterized in that, The extraction of time-domain and frequency-domain features from the acceleration data includes: The effective value, peak value, average value, variance, kurtosis, and skewness of the acceleration data are calculated to obtain the time-domain characteristics; The frequency domain characteristics are obtained by calculating the dominant frequency, total spectral energy, and spectral entropy of the acceleration data.

3. The method according to claim 1, characterized in that, The step of establishing a future health state model of the battery based on the vibration characteristics and the health state using a preset algorithm includes: Based on the vibration characteristics, a battery health status change rate model is established using a preset algorithm; Based on the stated health status and the battery health status change rate model, a future health status model for the battery is established.

4. The method according to claim 1, characterized in that, The method further includes: Acquire data samples of the target vehicle; the data samples include the target battery health status and target vibration characteristics; Based on the target vibration characteristics, the battery health status prediction of the target vehicle is determined by a predictive controller through a future health status model and a battery life objective function. Calculate the residual between the target battery health state and the predicted battery health state; If the residual is greater than a preset value, the data sample is added to the supplementary sample pool.

5. The method according to claim 1, characterized in that, The method further includes: Every preset number of days, or whenever the number of data samples in the supplementary sample pool reaches a preset number, the future health status model is retrained based on the data samples in the supplementary sample pool.

6. The method according to claim 1, characterized in that, Establishing battery operating constraints includes: Set the power range, temperature range, charge / discharge rate range, and remaining battery capacity range as battery operation constraints.

7. A battery life optimization device, characterized in that, The device includes: The acquisition module is used to acquire the vehicle's initial acceleration data, vehicle structural parameters, battery status parameters, and health status during driving; the battery status parameters include the vehicle battery's state of charge, power, and temperature. The extraction module is used to preprocess the initial acceleration data to obtain acceleration data; extract the time-domain and frequency-domain features of the acceleration data; and fuse the vehicle structural parameters, the time-domain features, the frequency-domain features, the power, the temperature, and the state of charge to obtain vibration features. A module is established to build a future health state model of the battery based on the vibration characteristics and the health state using a preset algorithm; the future health state model includes a battery health state change rate model; based on the battery health state change rate model, a battery life objective function and battery operation constraints are established; the battery life objective function minimizes the battery health state change rate by optimizing the battery's power and temperature; The prediction module is used to construct a prediction time period. Based on the vibration characteristics and the future health state model, the prediction controller predicts the trajectory of the battery's future health state change during the prediction time period. The future health state change trajectory is substituted into the battery life objective function. With the goal of minimizing the cumulative change in future battery health and combined with battery operation constraints, the prediction controller solves for the optimal power and temperature control sequence within the prediction time period. The control module is used to take the first power and temperature combination in the optimal power and temperature control sequence as the current optimal power and current optimal temperature; and to control the operating status of the vehicle through the vehicle controller based on the current optimal power and the current optimal temperature.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. 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 to 6.

10. A computer program product, comprising a computer program, 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 to 6.

Images