Bms control method and control system based on active identification and reverse compensation

Abstract

The application provides a BMS control method and control system based on active identification and reverse compensation, which is applied to a battery management system, and the method comprises the following steps: determining an SOC initial value based on a static voltage and an ambient temperature; calculating an SOC correction coefficient according to influencing factors and dynamically updating the SOC correction coefficient; fusing an ampere-hour integral method and a correction factor to complete SOC estimation; adopting an extended Kalman filtering algorithm to fuse voltage and current observation values to realize real-time prediction and correction of the SOC; combining a change rate of internal resistance and a capacity attenuation factor to obtain a current SOH value; extracting a characteristic curve and comparing a reference sample to realize health state identification, and then triggering reverse compensation of SOC estimation parameters according to the identification result, dynamically adjusting estimation parameters such as effective capacity and internal resistance, and improving SOC estimation accuracy under aging working conditions. The method is flexible in parameter updating, strong in SOC estimation robustness, and suitable for dynamic energy management and health management of a lithium iron phosphate battery pack in a complex working environment.

Description

Technical Field

[0001] This invention belongs to the field of battery management system technology, specifically relating to a BMS control method and control system based on active identification and reverse compensation. Background Technology

[0002] With the rapid development of electric vehicles, large-scale energy storage, and portable electronic devices, rechargeable batteries such as lithium-ion batteries serve as core power sources, making the accuracy and reliability of their management systems crucial. Accurate estimation of the core state parameters in the Battery Management System (BMS)—State of Charge (SOC) and State of Health (SOH)—is key to ensuring the safe, durable, and efficient operation of batteries.

[0003] Currently, the industry widely uses the ampere-hour integration method combined with the open-circuit voltage (OCV) method for State of Charge (SOC) estimation. The ampere-hour integration method estimates the change in charge by integrating the current; it is simple but suffers from accumulated error. The OCV method obtains an initial value by mapping the OCV to SOC after the battery has been idle for a sufficiently long time, but it cannot be directly applied under dynamic operating conditions. To correct the error of the ampere-hour integration, state observation algorithms such as the Extended Kalman Filter (EKF) have been introduced, using voltage feedback to correct the SOC, thus improving the estimation accuracy to some extent.

[0004] However, existing fusion methods still face several serious challenges: First, the usable capacity of a battery is not fixed but dynamically decays with environmental factors. Traditional SOC estimation models typically use a fixed nominal capacity or only perform coarse-grained compensation, resulting in a significant decrease in estimation accuracy throughout the battery's lifespan and under complex operating conditions. Second, SOH estimation is often independent of the SOC estimation loop, relying heavily on periodic capacity calibration or simple empirical models. This fails to reflect the battery's internal degradation dynamics in real time, causing the SOC estimation model to fail to adjust promptly when the battery's health changes, leading to systematic biases. For example, if a battery that has decayed to 80% of its original capacity is still estimated with 100% nominal capacity, the result will deviate significantly from the true value, potentially causing safety hazards such as overcharging and over-discharging. Summary of the Invention

[0005] In view of the above-mentioned problems in the prior art, the purpose of this invention is to provide a BMS control method based on active identification of voltage and current trajectories and reverse compensation of SOH, which can feed back the SOH changes to the core parameters of SOC estimation in real time and adaptively, and construct a dynamic closed-loop system to significantly improve the state estimation accuracy and reliability of BMS in all operating conditions and throughout the entire life cycle.

[0006] A BMS control method based on active voltage and current trajectory identification and SOH reverse compensation is characterized by the following steps: Step S1: When the battery is in a static state, collect the battery's open-circuit voltage and ambient temperature, and query a preset calibration table based on the open-circuit voltage and ambient temperature to obtain the battery's initial state of charge (SOC(0)). Step S2: Dynamically calculate the correction coefficient based on temperature, number of cycles and charge / discharge rate, and use the correction coefficient to calculate the current equivalent capacity Q of the battery; Step S3: Based on the initial state of charge (SOC) (0) and the current equivalent capacity (Q), perform real-time SOC estimation using the ampere-hour integration method; Step S4: Using an extended Kalman filter, the voltage and current observations of the battery are fused to correct the SOC estimated in step S3; Step S5: Estimate the state of health (SOH) of the battery and back-compensate the estimated SOH value into the SOC estimation model to form a closed-loop correction between SOC and SOH; Step S5 includes: Step S51: Extract key trajectory features from the battery's voltage and current time-series data to characterize the decline in battery health. Step S52: Input the extracted key trajectory features into the preset key trajectory feature-SOH mapping model to obtain the real-time SOH estimate. ; Step S53: Utilize the estimated SOH value Adjusting the current equivalent capacity Q yields the battery equivalent capacity after reverse compensation. And the equivalent capacity of the battery after reverse compensation Update the capacity parameters in the ampere-hour integral method and the extended Kalman filter to obtain the current battery SOC estimate based on SOH reverse compensation.

[0007] Preferably, the battery is in a static state under the following conditions: the change in battery current is less than the current set value and the static time continues to exceed the time threshold.

[0008] Preferably, the correction coefficient in step S2 includes: Temperature correction factor It is obtained by looking up a table of temperature and capacity, and is represented as the available capacity at the current temperature. usable capacity at standard temperature The ratio; Cyclic coefficient correction coefficient It is obtained by querying the battery aging degradation model or table, and is represented as the available capacity at the current cycle number. With the battery's factory rated capacity The ratio; Ratio Correction Factor It is obtained by querying the ratio and capacity relationship table, and is represented as the actual capacity under the current ratio I. Nominal capacity at 1C rate The ratio; The formula for calculating the current equivalent capacity Q is:

[0009] in, This refers to the battery's rated capacity.

[0010] Preferably, in step S3, the formula for calculating the real-time SOC using the ampere-hour integration method is as follows:

[0011] in, This refers to the battery's rated capacity.

[0012] Preferably, the key trajectory features extracted in step S51 include at least: Voltage change rate:

[0013] Discharge phase inflection point drift:

[0014] Pressure difference changes within the OC range: .

[0015] Preferably, the functional expression of the key trajectory features and the SOH mapping model in step S52 is as follows: .

[0016] Preferably, step S53 specifically includes: Based on the real-time SOH estimation results, obtain the battery equivalent capacity after reverse compensation:

[0017] Where Q represents the current equivalent capacity of the battery; : Indicates a reference health value; : Indicates the current estimated battery health status; The formula for estimating the current battery SOC based on SOH reverse compensation is: .

[0018] Another objective of this invention is to propose a BMS control system based on active voltage and current trajectory identification and SOH reverse compensation, comprising: The data acquisition module is used to acquire the open-circuit voltage and ambient temperature of the battery when the battery is in a static state, and to determine the initial state of charge (SOC) of the battery based on a preset calibration table. The capacity correction module is used to dynamically obtain the corresponding correction coefficients based on the current ambient temperature, number of cycles, and charge / discharge rate, by querying the relationship tables or models between temperature and capacity, rate and capacity, and cycle life and capacity, respectively, and then calculating the current equivalent capacity Q of the battery. The SOC estimation module is used to estimate the SOC in real time based on the initial SOC(0) and the current equivalent capacity Q using the ampere-hour integration method, and further calls the extended Kalman filter to fuse the current voltage and current observations to dynamically correct the SOC. The trajectory feature extraction module is used to extract key trajectory features characterizing the aging state of the battery from the voltage and current time series data of the battery. The trajectory features include, but are not limited to, voltage change rate, discharge stage inflection point drift, and constant voltage stage voltage difference change value. The SOH estimation module is used to input the extracted key trajectory features into the pre-trained mapping model between trajectory features and SOH to obtain the real-time estimated SOH value; and to feed back the real-time estimated SOH value to the SOC estimation module for reverse compensation to achieve dynamic correction. The control execution module is used to dynamically adjust the charge and discharge control, early warning strategy, and battery scheduling strategy in the battery management system based on the final estimated SOC and SOH states.

[0019] The beneficial effects of this invention are as follows: This BMS control method and control system based on active identification and reverse compensation establishes a capacity correction model by introducing multi-source correction parameters including temperature, rate, and cycle count. This enables rapid estimation of the current equivalent capacity based on a fixed functional relationship under different temperature and rate conditions, serving as the initial capacity input for SOC estimation, significantly improving the state adaptability of traditional BMS under non-standard operating conditions. The correction model obtains correction coefficients through an empirical mapping function, reasonably modeling the capacity decay law under environmental factors such as temperature and rate, thus improving the accuracy of SOC estimation in the early deployment stage.

[0020] By extracting voltage and current trajectory features and constructing a State of Health (SOH) estimation model, dynamic learning and identification of battery degradation states are achieved. Innovatively, trajectory features such as voltage change rate and discharge inflection point drift are extracted from dynamically generated voltage and current time-series data. These features can sensitively capture microscopic degradation phenomena such as loss of active materials and increase in internal resistance within the battery. Real-time, online estimation of SOH can be achieved without placing the battery in a static state or under specific testing conditions, which is beneficial for early fault warning and status monitoring.

[0021] Furthermore, through the SOH reverse compensation mechanism, the estimated SOH value is used to re-correct the equivalent capacity, making the SOC change corresponding to the unit current in the SOC estimation formula more consistent with the actual aging state, thus avoiding the problem of SOC overestimation caused by battery aging. This reverse compensation mechanism, using SOH as a bridge, establishes a dynamic feedback path between capacity degradation and SOC integration accuracy, and is an intelligent capacity calibration method that can adaptively adjust as the battery degrades.

[0022] This BMS control method and control system based on active identification and reverse compensation organically integrates multiple stages, including initial calibration, dynamic compensation, state filtering, feature recognition, mapping modeling, and reverse feedback, into a collaborative system. Each stage supports and enhances the others, collectively forming a hierarchical and highly responsive advanced battery state estimation algorithm architecture, providing a complete solution for high-precision and high-reliability battery management. Attached Figure Description

[0023] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0024] Example 1 like Figure 1 As shown, a BMS control method based on active identification of voltage and current trajectories and reverse compensation of SOH is proposed. By integrating environmental parameter correction and dynamic SOH estimation as dual paths, a battery equivalent capacity estimation mechanism is constructed, which achieves high accuracy, high adaptability and strong robustness of battery state estimation.

[0025] Includes the following steps: Step 1: Obtain initial SOC value When the battery is detected to be in a static state, the open-circuit voltage of the battery is collected. Given the ambient temperature T, the initial SOC value is obtained by consulting the two-dimensional mapping table of open-circuit voltage and temperature calibrated in the experiment. The function expression is: ,in, This is a table lookup function.

[0026] The conditions for the battery to be in a static state are: the change in battery current is less than the current setting value and the static time continues to exceed the time threshold.

[0027] For example: when (C is the battery's rated capacity) and the duration t>30s, the battery is considered to have reached the open circuit state, and the open circuit voltage can be accurately measured.

[0028] Step 2: Dynamically update the correction coefficients Considering the impact of factors such as temperature, cycle number, and charge / discharge rate on the actual battery capacity, a correction coefficient is introduced and dynamically updated to obtain the current equivalent capacity Q of the battery.

[0029] The correction factors include: Temperature correction factor It is used to correct for the effect of temperature on actual capacity and is defined as:

[0030] in, This represents the actual usable capacity of the battery at the current ambient temperature T, a value obtained through actual experimental calibration. The usable battery capacity at standard temperature (25°C) is obtained from the manufacturer's calibration table.

[0031] Specifically, the actual usable capacity of the battery at the current ambient temperature T. By setting multiple test temperature points and conducting repeated experiments, the capacity value is obtained by averaging the values ​​at each test point. Data fitting is then performed to form a corresponding fitting function, which facilitates the accurate determination of the battery's actual usable capacity based on the collected ambient temperature T. Numerical value.

[0032] Cyclic coefficient correction factor: It is used to correct capacity deviations caused by battery life degradation and is defined as:

[0033] in, This represents the usable capacity after N charge-discharge cycles, a value obtained from an aging degradation table or aging degradation model. The battery's factory-calibrated capacity.

[0034] Charge / discharge rate correction factor: This is used to correct for the impact of high-rate charge / discharge on capacity estimation. It is defined as:

[0035] in, This indicates the actual capacity when discharging at a current rate of I. This indicates the nominal capacity of the battery at a 1C rate.

[0036] The revised formula for calculating the equivalent battery capacity is as follows: ,in, This refers to the battery's rated capacity.

[0037] This is used to adjust the cumulative power consumption during the ampere-hour integration estimation process, thereby improving the adaptability and accuracy of SOC calculations throughout the entire life cycle, across multiple temperature zones, and at multiple rates.

[0038] Step 3: Preliminary estimation of SOC Based on the initial SOC value obtained in step 1 and the corrected battery equivalent capacity Q obtained in step 2, the real-time SOC is calculated using the ampere-hour integration method:

[0039] By integrating the charging and discharging current I over time and combining it with the capacity factor Q, the change in charge is estimated and added to the initial value to obtain the current State of Charge (SOC). Furthermore, since the capacity factor Q has been adjusted by a correction coefficient, the integration weight is increased to compensate for underestimation bias when the actual usable capacity decreases, which can effectively improve the accuracy of the current SOC estimate.

[0040] Step 4: Correct SOC using the EKF algorithm To improve the error accumulation and sensor drift problems of the ampere-hour integration method during long-term operation and further enhance the dynamic response accuracy of SOC estimation under complex operating conditions, an extended Kalman filter (EKF) is introduced as a nonlinear state estimation method based on the preliminary SOC estimation using the ampere-hour integration method, to perform dynamic prediction and observation correction of SOC.

[0041] The EKF algorithm uses SOC as the state variable, charging and discharging current I as the system input, and output voltage as a nonlinear observation to construct a prediction and observation fusion model.

[0042] Specifically, the battery's state of charge (SOC) is first modeled as a system state variable, and a dynamic SOC change model based on current input is established. Specifically, the battery's SOC at any given time... Predicted value It can be expressed as an ampere-hour integral:

[0043] The known SOC value at the previous time step; : The battery current collected at the current moment; Q: Current sampling period; Q: Battery equivalent capacity after correction based on temperature, number of cycles, and charge / discharge rate.

[0044] Then, using the battery's output voltage As an observation, a nonlinear observation model of the voltage is constructed, using the open-circuit voltage as an example. Mapping function between SOC As a voltage prediction benchmark, the predicted voltage is obtained by combining the internal resistance voltage drop and measurement noise. :

[0045] in, : The OCV–SOC curve function calibrated experimentally; The current equivalent internal resistance value can be obtained as a constant or through dynamic estimation. : The battery current collected at the current moment.

[0046] At this time, the actual voltage value is... With predicted voltage The difference between them constitutes the observation residual. The expression is:

[0047] Using this observation residual Multiplying the predicted SOC by the Kalman gain K corrects the SOC estimate, yielding the corrected SOC estimate. : .

[0048] Through the above EKF steps, the system can accurately correct the SOC under dynamic current disturbances and non-ideal observation conditions by combining the nonlinear observation model of voltage and the actual observation error, which significantly improves the estimation stability and response speed. It is especially suitable for BMS control scenarios with high-rate fluctuations and non-static states.

[0049] Step 5: SOH Estimation and Reverse Compensation Characteristic quantities representing the battery's health state, such as voltage change rate, discharge inflection point drift, and voltage difference within a specific range, are extracted from the voltage and current trajectories. A mapping model between key trajectory features and State of Health (SOH) is constructed using a BP neural network to achieve online estimation of the battery's SOH. The obtained SOH values ​​are used to inversely correct the capacity Q parameter, thereby indirectly affecting the SOC estimation and enabling the system to adapt to aging effects. The specific steps are as follows: Step 5.1 Trajectory Feature Extraction The system collects raw data sequences of voltage and current changes over time in real time, and extracts key trajectory features through methods such as differencing and fitting to characterize the degree of battery performance degradation.

[0050] Key trajectory features include: (1) Rate of voltage change: This represents the rate at which the voltage rises or falls over time. Internal polarization or capacitance loss can cause this value to drift.

[0051] (2) Discharge inflection point drift: : Indicates the offset of the voltage steep change inflection point on the SOC axis. For the current SOC, This is the corresponding reference SOC.

[0052] (3) Pressure difference changes within a specific OC range: This indicates the degree of voltage change of the battery during a period of near-constant current discharge. For example: It indicates the degree of voltage change of the battery as it discharges from 90% SOC to 70% SOC.

[0053] Step 5.2: Construct a mapping model between key trajectory features and SOH. A mapping model from extracted features to State of Health (SOH) was established using methods such as backpropagation (BP) neural networks. This model was trained with experimentally calibrated data and outputs real-time estimates of the state of health.

[0054] The functional expression for the key trajectory features and the SOH mapping model is as follows:

[0055] in, This can be a multiple linear regression model or a backpropagation (BP) neural network.

[0056] Step 5.3: SOH reverse compensation to SOC estimation model The real-time estimated SOH value is dynamically fed back to the SOC estimation process to achieve capacity adaptive adjustment and bidirectional coupling control based on the health status. This allows the SOC estimation model to follow the changes in the aging degree of the battery in real time, avoiding the systematic deviation that occurs in the later stage of aging in the traditional fixed capacity model, thereby constructing a dynamic closed-loop relationship between SOC and SOH.

[0057] Based on the real-time SOH estimation results, obtain the battery equivalent capacity after reverse compensation:

[0058] Where Q represents the current equivalent capacity of the battery; : Indicates a reference health value; This indicates the current estimated battery health status.

[0059] The formula for estimating the current battery SOC based on SOH reverse compensation is:

[0060] This compensation mechanism ensures that even if the battery capacity declines due to aging, the SOC estimate still accurately reflects the actual remaining capacity, avoiding the accumulation of systematic biases.

[0061] Example 2 A battery management system based on active voltage and current trajectory identification and reverse SOH compensation includes a data acquisition module, a capacity correction module, a SOC estimation module, a trajectory feature extraction module, an SOH estimation module, and a control execution module. The modules work together to construct a battery equivalent capacity estimation mechanism by integrating environmental parameter correction and dynamic SOH estimation, thereby achieving high accuracy, high adaptability, and strong robustness in battery state estimation.

[0062] The data acquisition module is used to collect the open-circuit voltage and ambient temperature of the battery when it is in a static state, and to determine the initial state of charge (SOC) of the battery based on a preset calibration table.

[0063] The capacity correction module is used to dynamically obtain the corresponding correction coefficients based on the current ambient temperature, number of cycles, and charge / discharge rate, by querying the relationship tables or models between temperature and capacity, rate and capacity, and cycle life and capacity, respectively, and then calculating the current equivalent capacity Q of the battery.

[0064] The SOC estimation module is used to estimate the SOC in real time based on the initial SOC(0) and the current equivalent capacity Q using the ampere-hour integration method, and further calls the extended Kalman filter to fuse the current voltage and current observations to dynamically correct the SOC.

[0065] The trajectory feature extraction module is used to extract key trajectory features characterizing the aging state of the battery from the voltage and current time series data of the battery. The trajectory features include, but are not limited to, voltage change rate, discharge stage inflection point drift, and constant voltage stage voltage difference change value.

[0066] The SOH estimation module is used to input the extracted key trajectory features into the pre-trained mapping model between trajectory features and SOH to obtain the real-time estimated SOH value; and to feed back the real-time estimated SOH value to the SOC estimation module for reverse compensation to achieve dynamic correction.

[0067] The control execution module is used to dynamically adjust the charge and discharge control, early warning strategy and battery scheduling strategy in the battery management system based on the final estimated SOC and SOH states.

[0068] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A BMS control method based on voltage-current trajectory active identification and SOH reverse compensation, characterized in that, Includes the following steps: Step S1: When the battery is in a static state, collect the battery's open-circuit voltage and ambient temperature, and query a preset calibration table based on the open-circuit voltage and ambient temperature to obtain the battery's initial state of charge (SOC(0)). Step S2: Dynamically calculate the correction coefficient based on temperature, number of cycles and charge / discharge rate, and use the correction coefficient to calculate the current equivalent capacity Q of the battery; The correction factor in step S2 includes: temperature correction coefficient obtained by looking up a temperature vs. capacity table, and is expressed as a ratio of the available capacity at the current temperature to the available capacity at a standard temperature the available capacity at the standard temperature​ Cycle coefficient correction factor obtained by querying a battery aging decay model or table, expressed as a ratio of the available capacity at the current cycle number to the battery's factory-calibrated capacity to the battery's factory-calibrated capacity ​ Ratio Correction Factor It is obtained by querying the ratio and capacity relationship table, and is represented as the actual capacity under the current ratio I. Nominal capacity at 1C rate The ratio; The formula for calculating the current equivalent capacity Q is: in, This refers to the battery's rated capacity. Step S3: Based on the initial state of charge (SOC) (0) and the current equivalent capacity (Q), perform real-time SOC estimation using the ampere-hour integration method; Step S4: Using an extended Kalman filter, the voltage and current observations of the battery are fused to correct the SOC estimated in step S3; Step S5: Estimate the state of health (SOH) of the battery and back-compensate the estimated SOH value into the SOC estimation model to form a closed-loop correction between SOC and SOH; Step S5 includes: Step S51: Extract key trajectory features from the battery's voltage and current time-series data to characterize the decline in battery health. Step S52: Input the extracted key trajectory features into the preset key trajectory feature-SOH mapping model to obtain the real-time SOH estimate. ; Step S53: Utilize the estimated SOH value Adjusting the current equivalent capacity Q yields the battery equivalent capacity after reverse compensation. And the equivalent capacity of the battery after reverse compensation Update the capacity parameters in the ampere-hour integral method and the extended Kalman filter to obtain the current battery SOC estimate based on SOH reverse compensation.

2. The BMS control method based on active voltage and current trajectory identification and SOH reverse compensation according to claim 1, characterized in that, The conditions for the battery to be in a static state are: the change in battery current is less than the current set value and the static time continues to exceed the time threshold.

3. The BMS control method based on active voltage and current trajectory identification and SOH reverse compensation according to claim 1, characterized in that, In step S3, the formula for calculating the real-time SOC using the ampere-hour integration method is as follows: in, This refers to the battery's rated capacity.

4. A BMS control system based on active voltage and current trajectory identification and SOH reverse compensation, used to implement the BMS control method based on active voltage and current trajectory identification and SOH reverse compensation as described in claim 1, characterized in that, include: The data acquisition module is used to acquire the open-circuit voltage and ambient temperature of the battery when the battery is in a static state, and to determine the initial state of charge (SOC) of the battery based on a preset calibration table. The capacity correction module is used to dynamically obtain the corresponding correction coefficients based on the current ambient temperature, number of cycles, and charge / discharge rate, by querying the relationship tables or models between temperature and capacity, rate and capacity, and cycle life and capacity, respectively, and then calculating the current equivalent capacity Q of the battery. The SOC estimation module is used to estimate the SOC in real time based on the initial SOC(0) and the current equivalent capacity Q using the ampere-hour integration method, and further calls the extended Kalman filter to fuse the current voltage and current observations to dynamically correct the SOC. The trajectory feature extraction module is used to extract key trajectory features characterizing the aging state of the battery from the voltage and current time series data of the battery. The key trajectory features include the voltage change rate, the inflection point drift during the discharge stage, and the voltage difference change value during the constant voltage stage. The SOH estimation module is used to input the extracted key trajectory features into the pre-trained mapping model between trajectory features and SOH to obtain the real-time estimated SOH value; and to feed back the real-time estimated SOH value to the SOC estimation module for reverse compensation to achieve dynamic correction. The control execution module is used to dynamically adjust the charge and discharge control, early warning strategy, and battery scheduling strategy in the battery management system based on the final estimated SOC and SOH states.

Images