Battery management method and device with automatic parameter configuration and sensor fault self-repair

Abstract

This invention provides a battery management method and apparatus with automatic parameter configuration and sensor fault self-repair, relating to the technical field of battery management. The method includes: real-time acquisition of battery operating signals from an electric vehicle using a sensor array, and preprocessing the battery operating signals to obtain a target battery operating signal; estimation of the battery's state of charge and health state based on a preset control model, using a battery equivalent circuit model and the target battery operating signal, to obtain a battery state estimation result, and determining the current model parameters of the battery equivalent circuit model based on the target battery operating signal; determining the target threshold and target parameters of the preset control model based on the current model parameters, and performing residual analysis on the battery state estimation result and the target battery operating signal to obtain the residual analysis result; and generating a target control command based on the target parameters, target threshold, and sensor fault information. This invention can significantly improve the safety of vehicle battery operation.

Description

Technical Field

[0001] This invention relates to the technical field of battery management, and in particular to a battery management method and apparatus for automatic parameter configuration and sensor fault self-repair. Background Technology

[0002] Battery management systems (BMS) are core components of electric vehicles and energy storage systems. Currently, mainstream BMS technologies acquire battery data periodically through data acquisition units, calculate the state of charge (SOC) and health status based on pre-calibrated fixed parameters and algorithm models, and trigger corresponding protection actions when monitored parameters exceed preset fixed thresholds. However, because the existing solutions use fixed model parameters, they cannot track time-varying characteristics as battery internal resistance increases with aging. This leads to an increase in SOC estimation errors as the battery ages. Furthermore, the system's fault judgment threshold is also a factory-fixed value and cannot adaptively adjust to changes in battery aging and temperature environment. In addition, when a sensor malfunctions, the system directly reports a fault and cuts off the high voltage, causing the vehicle to immediately lose power, affecting system availability and user experience. Therefore, existing battery management solutions lack dynamic adaptability. Their algorithm models and decision logic are static, and parameters are calibrated once and used for life. They cannot perceive and respond to changes in the battery's own state and the environment, thus affecting the safety of vehicle battery operation. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a battery management method and apparatus with automatic parameter configuration and sensor fault self-repair, which can significantly improve the safety of vehicle battery operation.

[0004] In a first aspect, embodiments of the present invention provide a battery management method for automatic parameter configuration and sensor fault self-repair. The method includes: acquiring battery operating signals of an electric vehicle in real time through a sensor set, and preprocessing the battery operating signals to obtain target battery operating signals, wherein the battery operating signals include voltage signals, current signals, and temperature signals; estimating the state of charge and health of the battery based on a preset control model, a battery equivalent circuit model, and the target battery operating signals to obtain battery state estimation results, and determining the current model parameters of the battery equivalent circuit model based on the target battery operating signals; determining the target threshold and target parameters of the preset control model based on the current model parameters, and performing residual analysis on the battery state estimation results and the target battery operating signals to obtain residual analysis results, and determining sensor fault information based on the residual analysis results; generating target control commands based on the target parameters, target threshold, and sensor fault information, and managing the battery by executing the target control commands.

[0005] In one implementation, the step of determining the current model parameters of the battery equivalent circuit model based on the target battery operating signal includes: using a recursive least squares algorithm with a forgetting factor to determine the current model parameters of the battery equivalent circuit model based on the current signal and voltage signal in the target battery operating signal.

[0006] In one implementation, the step of determining the target threshold and target parameters of the preset control model based on the current model parameters includes: adjusting the threshold information of the preset control model using the current model parameters to obtain the target threshold, and adjusting the parameters of the state of charge estimation model in the preset control model using the current model parameters to obtain the target parameters, wherein the target threshold includes: the maximum allowable discharge current, the charge / discharge cutoff voltage, and the equalization on threshold.

[0007] In one embodiment, the step of performing residual analysis on the battery state estimation result and the target battery operating signal to obtain the residual analysis result, and determining sensor fault information based on the residual analysis result, includes: performing residual calculation processing on the measured voltage value, measured current value, and measured temperature value in the target battery operating signal and the corresponding theoretical value in the battery state estimation result to obtain the residual analysis result; when the residual analysis result between any measured value and the corresponding theoretical value exceeds a preset residual threshold, it is determined that the sensor corresponding to the measured value has failed.

[0008] In one embodiment, after determining sensor fault information based on residual analysis results, the method includes: when a sensor fault is determined, performing fault reconstruction processing on the fault sensor based on the type of the faulty sensor and using measured data from the current healthy sensor, and determining the reconstruction value of the faulty sensor.

[0009] In one embodiment, the step of using measured data from current health sensors to perform fault reconstruction processing on faulty sensors and determine the reconstruction value of faulty sensors includes: when the faulty sensor is a voltage sensor, acquiring the measured voltage values ​​of two adjacent healthy cells and determining the arithmetic mean of the measured voltage values ​​as the reconstruction value of the faulty voltage sensor; when the faulty sensor is a current sensor, establishing a current prediction model using historical data collected before the fault, and obtaining the reconstruction value of the faulty current sensor by inputting the current vehicle speed, accelerator pedal opening, and brake pedal opening into the current prediction model; when the faulty sensor is a temperature sensor, inputting the current current value, coolant temperature, ambient temperature, and running time into a preset battery thermal network model to obtain the reconstruction value of the faulty temperature sensor.

[0010] In one implementation, the method further includes: sending a watchdog signal to the hardware watchdog at a preset period, and if the timeout occurs, rolling back all parameters to the factory safety baseline value.

[0011] Secondly, embodiments of the present invention also provide a battery management device with automatic parameter configuration and sensor fault self-repair. The device includes: a data preprocessing module, which collects battery operating signals of an electric vehicle in real time through a sensor set and preprocesses the battery operating signals to obtain target battery operating signals, wherein the battery operating signals include voltage signals, current signals, and temperature signals; a core control module, which estimates the state of charge and health of the battery based on a preset control model, a battery equivalent circuit model, and the target battery operating signals to obtain battery state estimation results, and determines the current model parameters of the battery equivalent circuit model based on the target battery operating signals; a self-repair module, which determines the target threshold and target parameters of the preset control model based on the current model parameters, and performs residual analysis on the battery state estimation results and the target battery operating signals to obtain residual analysis results, so as to determine sensor fault information based on the residual analysis results; and a battery management module, which generates target control commands based on the target parameters, target threshold, and sensor fault information, and manages the battery by executing the target control commands.

[0012] Thirdly, embodiments of the present invention also provide a server, including a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement any of the methods provided in the first aspect.

[0013] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement any of the methods provided in the first aspect.

[0014] The embodiments of the present invention bring the following beneficial effects: This invention provides a battery management method and apparatus for automatic parameter configuration and sensor fault self-repair. The method first collects real-time battery operating signals from an electric vehicle using a sensor set and preprocesses these signals to obtain a target battery operating signal. Then, using a preset control model, based on the battery equivalent circuit model and the target battery operating signal, it estimates the battery's state of charge and health to obtain a battery state estimation result. Based on the target battery operating signal, it determines the current model parameters of the battery equivalent circuit model and simultaneously determines the target threshold and target parameters of the preset control model. Residual analysis is performed between the battery state estimation result and the target battery operating signal to obtain residual analysis results, which are used to determine sensor fault information. Finally, based on the target parameters, target threshold, and sensor fault information, a target control command is generated, and the battery is managed by executing the target control command. This invention can significantly improve the safety of vehicle battery operation.

[0015] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of a battery management system with automatic parameter configuration and sensor fault self-repair provided in an embodiment of the present invention; Figure 2 This is a flowchart illustrating a battery management method for automatic parameter configuration and sensor fault self-repair provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of an adaptive parameter configuration method provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of a sensor fault reconstruction method provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of a battery management device with automatic parameter configuration and sensor fault self-repair provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of a server provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] The Battery Management System (BMS) is a core component of electric vehicles and energy storage systems. It is responsible for real-time monitoring of parameters such as voltage, current, and temperature of the battery pack, and performs functions such as state of charge (SOC) estimation, state of health (SOH) assessment, equalization control, thermal management, and fault protection. Currently, mainstream BMSs operate using fixed parameter models and threshold judgment methods. Its basic structure includes: a data acquisition unit (voltage / current / temperature sensors), a main control unit (MCU / DSP), an equalization unit, a communication unit, and an execution unit. The working principle is as follows: the data acquisition unit periodically acquires battery data; the main control unit calculates the SOC and SOH based on pre-calibrated lookup table parameters and fixed algorithm models (such as the ampere-hour integration method and open-circuit voltage method); and triggers corresponding protection actions when the monitored parameters exceed preset fixed thresholds.

[0021] To improve estimation accuracy, some advanced battery management systems (BMS) employ algorithms such as extended Kalman filtering for state estimation and introduce equivalent circuit models (e.g., first-order RC models and second-order RC models) to describe battery dynamic characteristics. The parameters of these models (e.g., ohmic internal resistance and polarization capacitance) are typically calibrated in a laboratory environment through mixed-pulse power characteristic testing and stored as fixed values ​​in the BMS. Existing BMS systems suffer from the following prominent problems in practical applications: First, fixed-parameter models struggle to adapt to parameter changes throughout the battery's entire lifecycle. During use, lithium-ion batteries undergo continuous internal electrochemical reactions, leading to increased ohmic internal resistance, decreased usable capacity, and altered polarization characteristics. For example, a new battery might have an internal resistance of 2 mΩ, which could increase to over 3 mΩ after 1000 cycles. Existing battery management systems (BMS) use fixed parameters calibrated in the laboratory, failing to track these time-varying characteristics. This results in a gradual increase in SOC estimation errors as the battery ages, from less than 3% initially to 5%-8% later.

[0022] Second, fixed threshold strategies cannot adapt to differences in operating conditions and aging levels. Existing BMS fault judgment thresholds (such as overvoltage protection threshold of 4.2V and overtemperature protection threshold of 55℃) are factory-fixed values. However, the safe operating range of batteries varies significantly at different temperatures (internal resistance increases at low temperatures, and the upper limit of allowable voltage should be appropriately lowered); batteries at different aging levels also have different tolerances to the same overcurrent. Fixed thresholds are either too conservative (sacrificing usable capacity) or too aggressive (posing safety hazards).

[0023] Third, sensor failures can lead to system malfunctions. Current BMS systems typically handle sensor failures by directly reporting the fault and cutting off the high voltage, causing the vehicle to immediately lose power. While this "either / or" approach ensures safety, it severely impacts system availability and user experience. For example, damage to a single voltage sampling channel can render the entire battery pack unusable, requiring in-store repair.

[0024] Fourth, manual calibration and maintenance are labor-intensive. To address these issues, existing solutions rely on engineers conducting extensive calibration experiments on test benches and in actual vehicles, establishing multi-dimensional parameter lookup tables (temperature-internal resistance tables, aging-capacity tables, etc.). This approach is time-consuming and labor-intensive, and it is difficult to cover all real-world usage scenarios.

[0025] The root cause of the above problems is that the existing BMS architecture lacks dynamic adaptability. Its algorithm model and decision-making logic are static. The parameters are calibrated once and used for life. Once the logic is solidified, it will not change. It cannot perceive and respond to changes in the battery's own state and the environment.

[0026] Based on this, the battery management method and apparatus for automatic parameter configuration and sensor fault self-repair provided by this invention can solve the problem of model parameter mismatch caused by aging and temperature changes throughout the battery's entire life cycle, and realize online identification and dynamic updating of battery model parameters; overcome the shortcomings of fixed threshold strategies in adapting to changes in operating conditions and aging levels, and realize adaptive adjustment of protection thresholds and control strategies; solve the problem of complete loss of BMS function caused by single-point sensor failure, and realize functional reconstruction and limp-mode operation under fault conditions; reduce manual parameter calibration and maintenance work throughout the battery's entire life cycle, and reduce the calibration and maintenance costs of the BMS.

[0027] To facilitate understanding of this embodiment, a detailed description of a battery management method for automatic parameter configuration and sensor fault self-repair disclosed in this embodiment of the invention will be provided first. This method is applied to a battery management system for automatic parameter configuration and sensor fault self-repair. To facilitate understanding of the battery management method for automatic parameter configuration and sensor fault self-repair, this embodiment of the invention provides a structural schematic diagram of a battery management system for automatic parameter configuration and sensor fault self-repair, as shown below. Figure 1 As shown, the battery management system includes the following components: (1) Sensing layer: includes voltage sensor array, current sensor and temperature sensor array, which are respectively connected to each cell and bus of the battery module to collect the voltage, current and temperature signals of the battery in real time.

[0028] (2) Data preprocessing layer: including signal conditioning circuit, analog-to-digital conversion unit and data cleaning module. The input of the signal conditioning circuit is connected to the output of each sensor to filter and amplify the original analog signal; the output of the signal conditioning circuit is connected to the input of the analog-to-digital conversion unit; the output of the analog-to-digital conversion unit is connected to the input of the data cleaning module to remove obvious abnormal data and noise.

[0029] (3) Core Control Layer: This layer includes a SOC estimation module, a SOH estimation module, a balance control module, a thermal management module, and a fault handling module. The input terminals of each module are connected to the output terminals of the data cleaning module to receive the processed sensor data; the modules interact with each other via an internal bus. This layer implements the basic functions of the BMS, and its algorithm model is implemented using configurable parameters.

[0030] (4) The adaptive and self-healing engine layer includes: Online parameter identifier: The input is connected to the output of the data cleaning module. It uses a recursive least squares algorithm with a forgetting factor to calculate the ohmic internal resistance R0, polarization internal resistance Rp and polarization capacitance Cp of the battery equivalent circuit model in real time.

[0031] Adaptive calibrator: The input is connected to the output of the online parameter identifier and the output of the SOH estimation module. Based on the identified internal resistance change and the SOH estimation result, it dynamically adjusts the control parameters of each module in the core control layer (including SOC estimation model parameters, equalization threshold, charge / discharge cutoff voltage, power limit value, etc.).

[0032] Sensor fault diagnostic tool: The input end is connected to the output end of the data cleaning module and the output end of each estimation module in the core control layer. It uses residual analysis to detect sensor faults, that is, it compares the measured value with the theoretical value estimated based on the model. When the residual continuously exceeds the threshold, the corresponding sensor fault is determined.

[0033] Virtual sensor reconstructor: The input end connects to the output end of the sensor fault diagnostic tool. It has multiple pre-set reconstruction algorithms (voltage reconstruction algorithm, current reconstruction algorithm, temperature reconstruction algorithm). When the fault diagnostic tool determines that a sensor is faulty, it automatically selects and activates the corresponding reconstruction algorithm according to the fault type, and uses the data of the healthy sensor and the battery model to estimate the current value of the faulty sensor.

[0034] (5) Decision and Execution Layer: This layer includes a central decision-making unit, contactor drive circuit, fan drive circuit, equalization execution circuit, and communication interface. The input of the central decision-making unit is connected to the output of the core control layer and the output of the virtual sensor reconfigurator, respectively. It integrates all information to make the final control decision and controls the actuator actions through each drive circuit. The communication interface is connected to the vehicle network or cloud platform for reporting fault information and receiving remote commands.

[0035] (6) Security monitoring layer: including hardware watchdog circuit and parameter confidence interval checking module. The hardware watchdog circuit is connected to each module of the adaptive and self-healing engine layer to monitor its operating status; the input of the parameter confidence interval checking module is connected to the output of the online parameter identifier to limit the identified parameters within a preset reasonable range to prevent the algorithm from diverging and causing control abnormalities.

[0036] During normal operation, the system collects battery voltage, current, and temperature data in real time. This data is input into the core control layer for routine SOC estimation and equalization control, and simultaneously into the newly added adaptive and self-healing engine layer.

[0037] The adaptive engine employs a recursive least squares algorithm with a forgetting factor to identify the ohmic internal resistance, polarization internal resistance, and polarization capacitance of the battery's equivalent circuit model in real time on an embedded platform. The identification results are used to correct model parameters in the SOC estimation algorithm and to assess changes in battery health status through long-term trend analysis. When parameter changes exceed preset thresholds, the SOH (State of Health estimation result) is automatically updated, and relevant control strategy thresholds (such as charging cutoff voltage and power limiting) are adjusted.

[0038] The self-healing engine performs real-time diagnostics on sensor signals, detecting sensor faults through two methods: analytical redundancy (residual analysis based on model estimates and measured values) and hardware redundancy (mutual verification among multiple sensors). When a specific sensor fault is detected, the system automatically switches to a backup virtual sensor mode: reconstructing the estimated value of the faulty sensor using data from healthy sensors and a battery model, maintaining basic BMS functions, and simultaneously generating a fault report and uploading it via the communication interface.

[0039] The sensor outputs of the battery module are connected in series with the battery data acquisition board of the sensing layer. After filtering and other processing, the data acquisition board outputs to the BMS main control board. The input of the adaptive and self-healing engine layer is based on the data from the sensing layer. The online identifier processes the data according to the battery model to obtain the ohmic internal resistance R0, polarization internal resistance Rp, and polarization capacitance Cp. The online parameter identifier outputs the identified parameters to the adaptive calibrator. Based on the identified internal resistance changes and SOH estimation results, the control parameters of each module in the core control layer (including SOC estimation model parameters, equalization threshold, charge / discharge cutoff voltage, power limit value, etc.) are dynamically adjusted. The adaptive and self-healing engine layer outputs automatic calibration parameters to the decision layer for control updates. The decision layer outputs control signal values ​​to the execution layer.

[0040] based on Figure 1 The diagram shows the structure of a battery management system with automatic parameter configuration and sensor fault self-repair. This invention provides a detailed description of the battery management method with automatic parameter configuration and sensor fault self-repair. (See also...) Figure 2 The diagram shows a battery management method for automatic parameter configuration and sensor fault self-repair. The method mainly includes the following steps S202 to S208: Step S202: The battery operation signal of the electric vehicle is collected in real time through the sensor set, and the battery operation signal is preprocessed to obtain the target battery operation signal, wherein the battery operation signal includes: voltage signal, current signal and temperature signal.

[0041] In one implementation, preprocessing refers to the process of converting the raw analog signal output by the sensor into a standardized digital signal that can be processed by subsequent digital algorithms. Specifically, this includes the following steps: First, the raw analog signal is filtered and amplified by a signal conditioning circuit. Filtering removes high-frequency noise and electromagnetic interference from the signal, while amplification increases the weak signal to a voltage range that the analog-to-digital converter (ADC) can recognize. Then, the conditioned analog signal is sent to the ADC unit, where a continuous analog signal is converted into discrete digital quantities at a preset sampling frequency. Finally, an outlier removal process is performed on the digital signal by a data cleaning module. This involves identifying and removing outliers that clearly exceed reasonable physical limits, such as voltage jumps to zero or data points far exceeding the rated voltage, as well as filtering out residual random noise. After this preprocessing, a clean, continuous, and stable target battery operating signal is obtained, which is used for subsequent battery state estimation and parameter identification processing.

[0042] Step S204: Using a preset control model, based on the battery equivalent circuit model and the target battery operating signal, the state of charge and health of the battery are estimated to obtain the battery state estimation result. Based on the target battery operating signal, the current model parameters of the battery equivalent circuit model are determined.

[0043] In one implementation, the preset control model refers to a set of algorithms built into the battery management system for performing core calculation and control functions, including a state of charge (SOC) estimation module and a state of health (SH) estimation module. SOC refers to the ratio of the battery's current remaining charge to its usable capacity when fully charged, reflecting how much electrical energy the battery can still provide. SH refers to the degree of degradation of the battery's current capacity relative to its factory-rated capacity, also expressed as a percentage, reflecting the degree of battery aging. The equivalent circuit model is a mathematical model in the battery management system used to describe the external electrical behavior characteristics of the battery. Specifically, it is an equivalent circuit structure constructed using resistive and capacitive elements, used to simulate the voltage response characteristics of a real battery during charging and discharging. In this invention, a first-order or second-order resistive-capacitive equivalent circuit model is used.

[0044] When performing state-of-charge (POC) estimation, the preset control model uses the current signal from the target battery's current operating signal as the input to the equivalent circuit model and the voltage signal as the output, iteratively solving the problem using an extended Kalman filter algorithm. This algorithm includes state equations and observation equations. The observation equations, based on the equivalent circuit model, describe the quantitative relationships between battery terminal voltage and current, POC, and model parameters. In each sampling period, the algorithm recursively solves for the optimal POC estimate through prediction and correction steps based on the current and voltage signals, obtaining the current POC estimation result. When performing health state estimation, the algorithm analyzes the changing trend of the ohmic resistance identification value output by the online parameter identifier over a long period, and combines this with capacity decay characteristics to comprehensively assess the current health state of the battery. The POC estimation result and the health state estimation result are combined and output as the final battery state estimation result.

[0045] In one implementation, a recursive least squares algorithm with a forgetting factor can be used to determine the current model parameters of the battery's equivalent circuit model based on the current and voltage signals in the target battery's operating signal. As the battery ages during use, parameters such as ohmic internal resistance, polarization internal resistance, and polarization capacitance in its equivalent circuit model change over time; the current model parameters are the actual values ​​of these parameters at the current moment. By using the recursive least squares algorithm with a forgetting factor, taking the current and voltage signals from the target battery's operating signal as input, the system calculates the values ​​of these model parameters in real time during each sampling period. This allows the system to accurately grasp the battery's current true electrical characteristics, providing a basis for subsequent parameter adjustments.

[0046] Step S206: Determine the target threshold and target parameters of the preset control model based on the current model parameters, and perform residual analysis on the battery state estimation results and the target battery operation signal to obtain the residual analysis results, so as to determine the sensor fault information based on the residual analysis results.

[0047] In one implementation, see Figure 3The diagram illustrates an adaptive parameter configuration method. It utilizes current model parameters to adjust threshold information of a preset control model to obtain a target threshold. Furthermore, it uses the current model parameters to adjust parameters of the state-of-charge estimation model within the preset control model to obtain target parameters. These target thresholds include: maximum allowable discharge current, charge / discharge cutoff voltage, and equalization activation threshold. The current model parameters are the parameters of the battery equivalent circuit model at the current moment, obtained in real-time through a recursive least squares algorithm with a forgetting factor. These parameters include ohmic internal resistance, polarization internal resistance, and polarization capacitance. The preset control model refers to the set of algorithms used in the battery management system to execute core control functions. Its internal control parameters and judgment thresholds are configurable, allowing for dynamic adjustment based on the battery state.

[0048] Specifically, the parameters of the state-of-charge (POC) estimation model in the preset control model are adjusted using the current model parameters. The POC estimation model employs an extended Kalman filter algorithm and contains an observation matrix. This matrix contains the model parameters of the equivalent circuit model, determining the quantitative relationship between voltage and current in the algorithm. The ohmic internal resistance identification value from the current model parameters is written into the corresponding parameter position in this observation matrix. This ensures that the POC estimation model establishes the observation equation based on the actual internal resistance value of the battery in each iteration, thereby achieving a consistent match between the model parameters and the current state of the battery.

[0049] Simultaneously, the threshold information of the preset control model is adjusted using the current model parameters. The threshold information includes the maximum allowable discharge current, charge / discharge cutoff voltage, and equalization activation threshold. The charge / discharge cutoff voltage is the highest voltage the battery can reach during charging and the lowest voltage it can drop to during discharging; the maximum allowable discharge current is the maximum current value allowed to be output by the battery in its current state; and the equalization activation threshold is the critical voltage difference value for activating the cell equalization function. The adjustment method is as follows: monitor the growth rate of the ohmic internal resistance identification value relative to the battery's initial internal resistance value; when a preset threshold is reached, adjust the charge / discharge cutoff voltage according to a preset mapping relationship; dynamically calculate the maximum allowable discharge current value based on the current temperature signal and the ohmic internal resistance value, according to the battery temperature rise constraint; when the health status estimation result is lower than the preset health threshold, tighten the equalization activation threshold from the initial voltage difference to a smaller voltage difference to strengthen the consistency management of aging batteries.

[0050] In another implementation, the measured voltage, current, and temperature values ​​from the target battery's operating signal are compared with their corresponding theoretical values ​​in the battery state estimation results using residual calculation. This yields residual analysis results. When the residual analysis result between any measured value and its corresponding theoretical value exceeds a preset residual threshold, the sensor corresponding to that measured value is determined to be faulty. The residual refers to the difference between the measured value of a sensor in the target battery's operating signal and the theoretical value of that physical quantity calculated based on the battery state estimation results. The measured voltage, current, and temperature values ​​from the target battery's operating signal are compared with their corresponding theoretical values ​​in the battery state estimation results using residual calculation. When the residual between any measured value and its corresponding theoretical value continuously exceeds a preset residual threshold, the sensor corresponding to that measured value is determined to be faulty, thus obtaining sensor fault information. This sensor fault information indicates which specific sensor has failed.

[0051] Step S208: Generate target control commands based on target parameters, target thresholds and sensor fault information, and manage the battery by executing the target control commands.

[0052] Specifically, target parameters refer to the state-of-charge (POC) estimation model parameters updated from the current model parameters. This allows the POC estimation model to perform subsequent calculations with model parameters that match the battery's current actual state, resulting in a more accurate POC estimate. Target thresholds refer to the maximum permissible discharge current, charge / discharge cutoff voltage, and equalization activation threshold, adjusted from the current model parameters, serving as new standards for subsequent battery protection decisions.

[0053] When sensor fault information indicates that there is no sensor fault, the decision execution layer directly uses the estimated state of charge and state of health output from the core control channel as input, combines the updated target parameters and target thresholds, and makes judgments according to preset control logic to generate corresponding control commands. The decision execution layer is the functional unit in the battery management system responsible for integrating all information and making the final control decision. It has pre-set decision rules based on input quantities such as state of charge, state of health, voltage, current, and temperature.

[0054] In one implementation, see Figure 4The diagram illustrates a sensor fault reconstruction method. When a sensor fault is determined, based on the type of the faulty sensor, the method utilizes the measured data from currently healthy sensors to perform fault reconstruction processing on the faulty sensor and determine its reconstruction value. When the faulty sensor is a voltage sensor, the measured voltage values ​​of two adjacent healthy cells are acquired, and the arithmetic mean of these measured voltage values ​​is determined as the reconstruction value of the faulty voltage sensor. When the faulty sensor is a current sensor, a current prediction model is established using historical data collected before the fault, and the reconstruction value of the faulty current sensor is obtained by inputting the current vehicle speed, accelerator pedal opening, and brake pedal opening into the current prediction model. When the faulty sensor is a temperature sensor, the current current value, coolant temperature, ambient temperature, and running time are input into a preset battery thermal network model to obtain the reconstruction value of the faulty temperature sensor.

[0055] At this point, the decision execution layer obtains the reconstructed value of the faulty sensor obtained from the fault reconstruction process. This reconstructed value replaces the measured value of the faulty sensor and serves as the current usable value for the physical quantity in subsequent control logic judgments. Since the reconstructed value is an estimated value calculated based on the type of faulty sensor and using the measured data of currently healthy sensors, it can reflect the true state of the physical quantity within a certain accuracy range. Therefore, the system can continue to maintain basic monitoring and protection functions even in the state of sensor failure, avoiding the loss of the entire system's functionality due to a single-point sensor failure.

[0056] After generating target control commands, the battery is managed by executing these commands. Target control commands include contactor on / off commands, power-limiting operation commands, equalization start commands, and thermal management commands. Contactor on / off commands control the connection and disconnection of the high-voltage circuit, cutting off the high voltage to protect the battery and personnel safety when a serious fault is detected. Power-limiting operation commands limit the battery's maximum output power, reducing power output to control temperature rise when the battery ages or experiences abnormal temperatures. Equalization start commands activate the equalization circuit when the voltage difference between cells exceeds the equalization activation threshold, reducing inconsistencies between cells. Thermal management commands control the start / stop and speed of thermal management actuators such as fans and water pumps, ensuring the battery operates within a suitable temperature range. These actuators execute corresponding actions based on the received target control commands, achieving battery charge / discharge management, safety protection, and performance optimization.

[0057] The battery management method with automatic parameter configuration and sensor fault self-repair provided in this invention can improve the accuracy of state of charge estimation. Existing technologies using fixed model parameters cannot track parameter changes caused by battery aging, and the estimation error gradually increases with usage time. In contrast, this invention identifies the current model parameters online and updates the state of charge estimation model in real time, ensuring the model always matches the current battery state and significantly reducing estimation errors throughout the entire battery lifespan.

[0058] Furthermore, this invention can improve system availability. Existing technologies directly cut off high voltage when a sensor fails, causing the vehicle to become inoperable. In contrast, this invention detects sensor faults through residual analysis and reconstructs fault values ​​based on the fault type using data from healthy sensors to replace the measured values, enabling the system to maintain basic functionality even when a single sensor fails.

[0059] Finally, this invention can reduce manual maintenance costs and enhance system safety. Existing technologies rely on manual calibration to solve parameter deviation problems and require periodic return to the factory for calibration. In contrast, this invention achieves automatic parameter configuration throughout the entire life cycle through online parameter identification and automatic adjustment of control parameters, reducing the frequency of manual maintenance. The fixed thresholds of existing technologies cannot be adjusted with battery aging, resulting in false protection or untimely protection. In contrast, this invention dynamically adjusts the charge / discharge cutoff voltage, maximum allowable discharge current, and equalization threshold based on real-time identified internal resistance change trends and health status, ensuring that the protection boundary always matches the current state of the battery.

[0060] In practical applications, this invention is applied to the power battery management system of a pure electric vehicle. The battery pack uses ternary lithium batteries with a rated voltage of 350V and a capacity of 100Ah, consisting of 96 cells connected in series.

[0061] 1. Hardware configuration: Main control chip: The Infineon TC275 tri-core MCU is used, where Core0 runs the core control layer tasks, Core1 runs the adaptive and self-healing engine layer tasks, and Core2 runs the security monitoring and communication tasks, so as to achieve functional isolation and real-time guarantee.

[0062] Sensor configuration: Voltage sampling: The LTC6813 acquisition chip is used, which supports the acquisition of up to 18 series-connected cells. A total of 6 chips are used, with each chip responsible for 16 cells. The sampling accuracy is ±1mV. Current sampling: The LEMHASS600-S Hall current sensor is used, with a range of ±600A and an accuracy of ±0.5%. Temperature sampling: The NTC thermistor is used, with one temperature point arranged for every 4 cells, for a total of 24 temperature sensors.

[0063] Actuator: High-voltage contactors: one each for main positive, main negative, and pre-charge contactors; Active balancing circuit: each cell is equipped with a bidirectional flyback balancing circuit with a maximum balancing current of 1A; Thermal management: control interfaces for water pumps, fans, and PTC heaters.

[0064] Communication interfaces: CAN2.0B interface connects to the vehicle controller, and 4G wireless communication module connects to the cloud platform.

[0065] 2. Software Implementation: (1) Equivalent circuit model and online parameter identification. In this embodiment, a second-order RC equivalent circuit model is used to describe the dynamic characteristics of the battery. The model equation is: U L =U {OCV}(SOC) -iR0-U1-U2 Where U1 is the first RC network of the second-order RC equivalent circuit model, and U2 is the second RC network of the second-order RC equivalent circuit model. L U is the battery terminal voltage. {OCV}(SOC) Let R0 be the open-circuit voltage and R0 be the internal resistance in ohms. U1 and U2 satisfy the following: dU1 / dt = -U1 / (R1C1) + i / C1 dU2 / dt = -U2 / (R2C2) + i / C2 Where dU1 / dt is the rate of change of U1 with time, dU2 / dt is the rate of change of U2 with time, R1 is the first polarization internal resistance, R2 is the first polarization internal resistance, C1 is the first polarization capacitor, C2 is the second polarization capacitor, and i is the current.

[0066] The online parameter identifier uses recursive least squares with a forgetting factor (FFRLS) to estimate the parameter vector θ=[R0,R1,C1,R2,C2]^T in real time. The forgetting factor λ is set to 0.98, which gives the algorithm higher weight to recent data and can track the slow changes caused by battery aging. The algorithm executes once every 100ms, implemented in Core1, and the test results show that a single calculation takes about 3ms, which meets the real-time requirements.

[0067] (2) Adaptive SOC estimation. The SOC estimation module uses the Extended Kalman Filter (EKF) algorithm. In traditional methods, the state equation and observation equation parameters of EKF are fixed; in this embodiment, the adaptive calibrator writes the latest estimated R0, R1, C1, R2, and C2 from the online parameter identifier into the observation matrix of EKF in real time, so that the SOC estimation model always matches the current state of the battery.

[0068] Furthermore, when the vehicle is stationary for an extended period (≥2 hours), the system automatically records the open-circuit voltage U_OCV for calibrating the U_OCV-SOC curve. Combined with the online identified R0, the OCV curve under different aging conditions can be dynamically corrected, further improving the accuracy of SOC estimation.

[0069] (3) Adaptive threshold adjustment. The adaptive calibrator dynamically adjusts the following thresholds based on the online identified internal resistance R0 and SOH to estimate the health status of the module output: charging cut-off voltage: initial value 4.20V, when R0 increases by more than 20%, it is linearly reduced to 4.15V (to prevent overcharging of aging batteries); discharging cut-off voltage: initial value 2.80V, when R0 increases by more than 30%, it is linearly increased to 3.00V (to prevent over-discharging of aging batteries); maximum allowable discharge current: initial value 300A, dynamically calculated based on real-time temperature and internal resistance to ensure that the temperature rise does not exceed the safety limit; equalization activation threshold: initial value 20mV, when SOH is below 80%, it is tightened to 10mV (to strengthen the consistency management of aging batteries).

[0070] (4) Sensor fault self-repair. The sensor fault diagnostic tool uses the following methods to detect faults: Voltage sensor failure: Calculate the voltage difference between adjacent cells ΔV(i) = |V(i) - V(i-1)|. If ΔV is consistently greater than 100mV (normally it should be <30mV) and other adjacent differences are normal, the voltage sensor of the i-th cell is determined to be faulty.

[0071] Current sensor failure: Compare the measured current I_meas with the current I_est estimated based on motor torque (obtained from the vehicle controller). If the residual |I_meas-I_est| is consistently greater than 20A, the current sensor is determined to be faulty.

[0072] Temperature sensor malfunction: If the temperature change rate at a certain point is abnormal (>5℃ / s) or the temperature difference with other points exceeds 20℃, the temperature sensor is considered malfunctioning.

[0073] When a fault is detected, the virtual sensor reconstructor automatically activates the reconstruction algorithm: Voltage sensor fault reconstruction: using adjacent cell voltage interpolation method. V_fault(t)=(V(i-1,t)+V(i+1,t)) / 2 At the same time, active balancing for this cell is disabled, and the cell is marked as requiring enhanced monitoring.

[0074] Current sensor fault reconstruction: Switch to voltage prediction and ampere-hour integration mode. First, a simple current prediction model is trained using normal current data from the 24 hours prior to the fault. The inputs are vehicle speed, accelerator pedal opening, and brake pedal opening, and the output is the predicted current. After the fault, the predicted current is used to estimate the SOC change by ampere-hour integration. Each time the vehicle is stationary, the SOC is forcibly calibrated using the open-circuit voltage method to correct the accumulated error.

[0075] Temperature sensor fault reconstruction: estimation based on thermal model lookup table. T_fault=f(I,T_coolant,T_amb,t) The system has a built-in battery thermal network model. Input the current current, coolant temperature, ambient temperature and running time, look up the table to output the estimated temperature, and add a 5℃ safety margin for thermal management control.

[0076] After reconfiguration activation, the system reports fault codes via CAN, and stores the fault type, time, and data fragments before and after reconfiguration locally, and uploads them to the cloud platform via the 4G module for subsequent algorithm optimization.

[0077] (5) Safety Redundancy Mechanism. To ensure system safety in the event of an adaptive engine malfunction, the following redundancy mechanism is implemented: The hardware watchdog sends a feed signal to the hardware watchdog at a preset period. If the timeout occurs, the parameters will be rolled back to the factory safety baseline value. Core1 needs to feed the watchdog once every 50ms. If the watchdog is not fed within the timeout, the system will automatically roll back the core control layer parameters to the factory safety baseline value and light up the fault light to indicate maintenance.

[0078] Parameter confidence interval: The online identification of R0 is limited to the interval [0.5R0_init, 3R0_init]. When the identification value exceeds this range, the previous valid value is used and the flag is set.

[0079] Degradation switching: When the adaptive engine outputs abnormalities 3 times in a row (such as SOC jump > 10%), the system will force a switch back to the traditional lookup table mode within 10ms to ensure that basic functions are not interrupted.

[0080] 3. Actual vehicle verification results.

[0081] This embodiment involved a 6-month road test on 10 test vehicles, accumulating a total mileage of over 500,000 kilometers. The test results are as follows: SOC estimation accuracy: Compared with reference battery tester data, the average absolute error is 2.1% and the maximum error is 3.8% over the entire life cycle, which is better than the traditional method's 4.5% and 7.2%.

[0082] Sensor fault simulation: The voltage sampling line of the 20th cell was manually disconnected. The system detected the fault within 2 seconds and switched to reconfiguration mode. The vehicle maintained limited power driving (maximum speed 60km / h) and successfully drove to the repair point (50km away). During this period, the SOC estimation error was <5%.

[0083] Aging adaptability: Compared with test vehicles that have driven 80,000 kilometers (SOH approximately 85%), the SOC estimation error was 2.8% in adaptive mode and 5.6% in fixed parameter mode; the full charge capacity was 3.2% higher than that in fixed parameter mode (due to more accurate dynamic adjustment of cutoff voltage).

[0084] Fault self-repair success rate: A total of 236 sensor faults were simulated, of which 228 successfully activated reconstruction and maintained system operation, with a success rate of 96.6%; the 8 failures were mainly scenarios of simultaneous failure of multiple sensors.

[0085] Furthermore, for resource-constrained embedded platforms, a first-order RC model can be used instead of a second-order RC model to reduce computational load; for applications with functional safety requirements of ISO26262ASILC / D, redundant MCUs and heterogeneous implementations can be added to this architecture; for fixed scenarios such as energy storage power stations, a global optimization layer based on a cloud platform can be added to achieve coordinated adaptation across battery clusters.

[0086] In summary, the present invention has the following beneficial effects: First, it significantly improves the stability of SOC estimation accuracy throughout the entire battery lifecycle. This invention uses online parameter identification technology to track changes in battery internal resistance and polarization characteristics in real time, ensuring the equivalent circuit model always matches the battery's current state. Simulation tests show that throughout the battery's entire lifecycle (initial - 20% capacity decay), the SOC estimation error of this invention can be controlled within 3%, while traditional fixed-parameter methods can reach 5%-8% error in the later stages of battery aging. This means users can obtain more accurate range predictions, reducing range anxiety.

[0087] Second, it significantly improves system availability in the event of sensor failure. The self-healing module of this invention can automatically switch to virtual sensor mode in the event of a single-point sensor failure, maintaining the basic functions of the BMS. Taking a typical single-cell voltage sensor failure as an example: a traditional BMS would immediately cut off the high voltage, rendering the vehicle immobile; the system of this invention can switch to adjacent cell voltage estimation mode, supporting the vehicle to travel to the repair point in a power-limited mode (limp-down mode), increasing system availability from 0% to over 95%. Combining research findings from the University of Warwick, the improved recursive least squares algorithm can simultaneously track electrical and thermal parameters, further reducing reliance on the sensor network.

[0088] Third, it reduces the manual maintenance costs throughout the battery's lifecycle. Traditional BMS systems often require battery recalibration at the factory or in a repair shop as the battery ages. This invention enables online self-learning and adaptive adjustment of parameters, allowing the battery to become its own "calibration engineer," reducing the workload of parameter maintenance and calibration throughout its lifecycle by more than 70%. Related research from Hefei University of Technology also shows that adaptive BMS can achieve optimized control throughout the entire lifecycle through parameter self-adjustment.

[0089] Fourth, it enhances the system's safety and robustness. The adaptive calibrator of this invention can dynamically adjust the protection threshold based on the battery's current aging state and temperature conditions. For example, it automatically lowers the charging cut-off voltage in low-temperature environments to prevent lithium plating; and automatically widens the voltage protection window in the later stages of aging to avoid frequent false protection. Combined with engineering-adaptive electrochemical modeling technology, it enables early detection and isolation of minor faults such as micro-overcharging. Compared to fixed thresholds, this dynamic threshold strategy ensures both a safety margin and maximizes the battery's usable capacity.

[0090] Fifth, it reduces the BMS development cycle and calibration costs. Because the system possesses self-learning and adaptive capabilities, OEMs no longer need to conduct extensive full-condition calibration experiments for each vehicle model and each battery type; they only need to verify key safety boundaries. This is expected to shorten the BMS development cycle by 20% and reduce calibration testing costs by more than 30%.

[0091] Regarding the battery management method for automatic parameter configuration and sensor fault self-repair provided in the foregoing embodiments, this invention provides a battery management device for automatic parameter configuration and sensor fault self-repair. See [link to relevant documentation]. Figure 5 The diagram shows a battery management device with automatic parameter configuration and sensor fault self-repair capabilities. The device includes the following components: The data preprocessing module 502 collects the battery operation signals of the electric vehicle in real time through the sensor set, and preprocesses the battery operation signals to obtain the target battery operation signals, wherein the battery operation signals include: voltage signals, current signals and temperature signals; The core control module 504 estimates the state of charge and health of the battery based on the battery equivalent circuit model and the target battery operating signal by using a preset control model, and obtains the battery state estimation result. It also determines the current model parameters of the battery equivalent circuit model based on the target battery operating signal. The self-repair module 506 determines the target threshold and target parameters of the preset control model based on the current model parameters, and performs residual analysis on the battery state estimation results and the target battery operating signal to obtain the residual analysis results, so as to determine the sensor fault information based on the residual analysis results; The battery management module 508 generates target control commands based on target parameters, target thresholds, and sensor fault information, and manages the battery by executing the target control commands.

[0092] The battery management device with automatic parameter configuration and sensor fault self-repair provided in this application embodiment can significantly improve the safety of vehicle battery operation.

[0093] In one embodiment, when performing the step of determining the current model parameters of the battery equivalent circuit model based on the target battery operating signal, the core control module 504 is further configured to: use a recursive least squares algorithm with a forgetting factor to determine the current model parameters of the battery equivalent circuit model based on the current signal and voltage signal in the target battery operating signal.

[0094] In one embodiment, when performing the step of determining the target threshold and target parameters of the preset control model based on the current model parameters, the self-repair module 506 is further configured to: adjust the threshold information of the preset control model using the current model parameters to obtain the target threshold, and adjust the parameters of the state of charge estimation model in the preset control model using the current model parameters to obtain the target parameters, wherein the target threshold includes: the maximum allowable discharge current, the charge and discharge cutoff voltage, and the equalization start threshold.

[0095] In one embodiment, when performing residual analysis on the battery state estimation result and the target battery operating signal to obtain the residual analysis result, and determining sensor fault information based on the residual analysis result, the self-repair module 506 is further configured to: perform residual calculation processing on the measured voltage value, measured current value, and measured temperature value in the target battery operating signal and the corresponding theoretical value in the battery state estimation result to obtain the residual analysis result; when the residual analysis result between any measured value and the corresponding theoretical value exceeds a preset residual threshold, it is determined that the sensor corresponding to the measured value has failed.

[0096] In one embodiment, after performing the step of determining sensor fault information based on residual analysis results, the self-repair module 506 is further configured to: when it is determined that a sensor has failed, based on the type of the faulty sensor, use the measured data of the current healthy sensor to perform fault reconstruction processing on the faulty sensor and determine the reconstruction value of the faulty sensor.

[0097] In one embodiment, when performing the step of using the measured data of the current health sensor to reconstruct the faulty sensor and determine the reconstruction value of the faulty sensor, the self-repair module 506 is further configured to: when the faulty sensor is a voltage sensor, acquire the measured voltage values ​​of two adjacent healthy cells, and determine the reconstructed value of the faulty voltage sensor by the arithmetic mean of the measured voltage values; when the faulty sensor is a current sensor, establish a current prediction model using historical data collected before the fault, and obtain the reconstructed value of the faulty current sensor by inputting the current vehicle speed, accelerator pedal opening, and brake pedal opening into the current prediction model; when the faulty sensor is a temperature sensor, input the current current value, coolant temperature, ambient temperature, and running time into a preset battery thermal network model to obtain the reconstructed value of the faulty temperature sensor.

[0098] In one embodiment, the self-healing module 506 is further configured to: send a watchdog signal to the hardware watchdog at a preset period, and if the timeout occurs, roll back each parameter to the factory safety baseline value.

[0099] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0100] This invention provides a server, specifically, the server includes a processor and a storage device; the storage device stores a computer program, which, when run by the processor, executes the method described in any of the above embodiments.

[0101] Figure 6 This is a schematic diagram of the structure of a server provided in an embodiment of the present invention. The server 100 includes: a processor 60, a memory 61, a bus 62, and a communication interface 63. The processor 60, the communication interface 63, and the memory 61 are connected through the bus 62. The processor 60 is used to execute executable modules, such as computer programs, stored in the memory 61.

[0102] The memory 61 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 63 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0103] Bus 62 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0104] The memory 61 is used to store programs. After receiving an execution instruction, the processor 60 executes the program. The method executed by the device for defining the flow process disclosed in any of the foregoing embodiments of the present invention can be applied to the processor 60 or implemented by the processor 60.

[0105] Processor 60 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 60 or by instructions in software form. Processor 60 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 61. Processor 60 reads the information in memory 61 and, in conjunction with its hardware, completes the steps of the above method.

[0106] The computer program product of the readable storage medium provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For specific implementation, please refer to the foregoing method embodiments, which will not be repeated here.

[0107] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium 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 described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0108] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A battery management method with automatic parameter configuration and sensor fault self-repair, characterized in that, The method includes: The battery operating signals of the electric vehicle are collected in real time by a set of sensors, and the battery operating signals are preprocessed to obtain the target battery operating signals, wherein the battery operating signals include: voltage signals, current signals and temperature signals; By using a preset control model, based on the battery equivalent circuit model and the target battery operating signal, the state of charge and state of health of the battery are estimated to obtain the battery state estimation result. Based on the target battery operating signal, the current model parameters of the battery equivalent circuit model are determined. Based on the current model parameters, the target threshold and target parameters of the preset control model are determined, and residual analysis is performed on the battery state estimation results and the target battery operating signal to obtain residual analysis results, so as to determine sensor fault information based on the residual analysis results; Based on the target parameters, the target threshold, and the sensor fault information, a target control command is generated, and the battery is managed by executing the target control command. The step of determining the target threshold and target parameters of the preset control model based on the current model parameters includes: adjusting the threshold information of the preset control model using the current model parameters to obtain the target threshold, and adjusting the parameters of the state of charge estimation model in the preset control model using the current model parameters to obtain the target parameters, wherein the target threshold includes: maximum allowable discharge current, charge / discharge cutoff voltage and equalization on threshold; The step after determining the sensor fault information based on the residual analysis results includes: when it is determined that a sensor has failed, based on the type of the faulty sensor, using the measured data of the current healthy sensor, performing fault reconstruction processing on the faulty sensor, and determining the reconstruction value of the faulty sensor. The step of using measured data from current health sensors to perform fault reconstruction processing on the faulty sensor and determine the reconstruction value of the faulty sensor includes: when the faulty sensor is a voltage sensor, acquiring the measured voltage values ​​of two adjacent healthy cells and determining the arithmetic mean of the measured voltage values ​​as the reconstruction value of the faulty voltage sensor; when the faulty sensor is a current sensor, establishing a current prediction model using historical data collected before the fault, and obtaining the reconstruction value of the faulty current sensor by inputting the current vehicle speed, accelerator pedal opening, and brake pedal opening into the current prediction model; when the faulty sensor is a temperature sensor, inputting the current current value, coolant temperature, ambient temperature, and running time into a preset battery thermal network model to obtain the reconstruction value of the faulty temperature sensor.

2. The battery management method for automatic parameter configuration and sensor fault self-repair according to claim 1, characterized in that, The step of determining the current model parameters of the battery equivalent circuit model based on the target battery operating signal includes: Using a recursive least squares algorithm with a forgetting factor, the current model parameters of the battery equivalent circuit model are determined based on the current and voltage signals in the target battery's operating signal.

3. The battery management method for automatic parameter configuration and sensor fault self-repair according to claim 1, characterized in that, The step of performing residual analysis on the battery state estimation result and the target battery operating signal to obtain residual analysis results, and determining sensor fault information based on the residual analysis results, includes: The measured values ​​of voltage, current, and temperature in the target battery operating signal are compared with the corresponding theoretical values ​​in the battery state estimation results to perform residual calculations, thereby obtaining the residual analysis results. When the residual analysis result between any measured value and the corresponding theoretical value exceeds a preset residual threshold, it is determined that the sensor corresponding to the measured value has malfunctioned.

4. The battery management method for automatic parameter configuration and sensor fault self-repair according to claim 1, characterized in that, The method further includes: The watchdog timer sends a feed signal to the hardware watchdog at a preset period. If the timeout occurs, all parameters are rolled back to the factory safety baseline value.

5. A battery management device with automatic parameter configuration and sensor fault self-repair, characterized in that, The device includes: The data preprocessing module collects the battery operation signals of the electric vehicle in real time through a set of sensors, and preprocesses the battery operation signals to obtain the target battery operation signals, wherein the battery operation signals include: voltage signals, current signals and temperature signals; The core control module estimates the state of charge and state of health of the battery based on the battery equivalent circuit model and the target battery operating signal by using a preset control model, thereby obtaining the battery state estimation result. It also determines the current model parameters of the battery equivalent circuit model based on the target battery operating signal. The self-repair module determines the target threshold and target parameters of the preset control model based on the current model parameters, and performs residual analysis on the battery state estimation results and the target battery operating signal to obtain residual analysis results, so as to determine sensor fault information based on the residual analysis results; The battery management module generates target control commands based on the target parameters, the target threshold, and the sensor fault information, and manages the battery by executing the target control commands. The step of determining the target threshold and target parameters of the preset control model based on the current model parameters includes: adjusting the threshold information of the preset control model using the current model parameters to obtain the target threshold, and adjusting the parameters of the state of charge estimation model in the preset control model using the current model parameters to obtain the target parameters, wherein the target threshold includes: maximum allowable discharge current, charge / discharge cutoff voltage and equalization on threshold; The step after determining the sensor fault information based on the residual analysis results includes: when it is determined that a sensor has failed, based on the type of the faulty sensor, using the measured data of the current healthy sensor, performing fault reconstruction processing on the faulty sensor, and determining the reconstruction value of the faulty sensor. The step of using measured data from current health sensors to perform fault reconstruction processing on the faulty sensor and determine the reconstruction value of the faulty sensor includes: when the faulty sensor is a voltage sensor, acquiring the measured voltage values ​​of two adjacent healthy cells and determining the arithmetic mean of the measured voltage values ​​as the reconstruction value of the faulty voltage sensor; when the faulty sensor is a current sensor, establishing a current prediction model using historical data collected before the fault, and obtaining the reconstruction value of the faulty current sensor by inputting the current vehicle speed, accelerator pedal opening, and brake pedal opening into the current prediction model; when the faulty sensor is a temperature sensor, inputting the current current value, coolant temperature, ambient temperature, and running time into a preset battery thermal network model to obtain the reconstruction value of the faulty temperature sensor.

6. A server, characterized in that, The method includes a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement the method of any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions that, when invoked and executed by a processor, cause the processor to perform the method described in any one of claims 1 to 4.

Images