Lithium battery dynamic charging and discharging strategy optimization method based on multi-dimensional parameter fusion
By integrating Kalman filtering and particle swarm optimization algorithms, multi-dimensional parameter coordination of lithium battery charging and discharging strategies is achieved, solving the problem of difficult parameter interaction in existing technologies and improving the stability and safety of lithium batteries in dynamic environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 四川文理学院
- Filing Date
- 2025-12-08
- Publication Date
- 2026-04-21
AI Technical Summary
Existing lithium battery charge and discharge management methods cannot fully consider the interaction of multiple parameters, resulting in insufficient control accuracy in dynamic environments. In particular, under high load or extreme conditions, strategy misalignment is prone to occur, affecting battery stability and safety.
By collecting charging and discharging data of lithium batteries, noise suppression and state estimation are performed using the Kalman filter algorithm. The internal resistance change rate and current stability index are fused together, and the optimal current adjustment scheme is searched using the particle swarm optimization algorithm. The feedback loop for dynamic change coordination is determined, and coordination commands are output to ensure battery safety and stability.
It achieves adaptive modifications for lithium battery charging and discharging scenarios, effectively reduces safety risks, improves charging efficiency and battery stability, and provides a reliable battery management solution.
Smart Images
Figure CN121906752A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of battery charge and discharge management technology, and in particular to a method for optimizing dynamic charge and discharge strategies for lithium batteries based on multi-dimensional parameter fusion. Background Technology
[0002] Lithium-ion batteries, as a core technology in modern energy storage, are widely used in electric vehicles, portable devices, and new energy systems. Their performance and lifespan directly affect the stability and safety of these devices. Optimizing the charging and discharging strategies of lithium-ion batteries to improve their efficiency and extend their lifespan has become one of the key research directions in this field.
[0003] However, many current charge and discharge management methods often fail to comprehensively consider the various influencing factors during battery operation when facing complex usage scenarios. These methods typically focus on only one or two key parameters, neglecting the interactions between different parameters. This leads to insufficient control accuracy in dynamic environments, especially when battery state changes rapidly, making strategy misalignment a common problem. This limitation results in poor battery performance under high loads or extreme conditions, affecting the overall system reliability.
[0004] Against this backdrop, optimizing lithium battery charging and discharging strategies faces significant technical challenges. The core issue lies in effectively integrating multiple key parameters, such as voltage, current, and temperature, to achieve a comprehensive understanding of the battery's state. These parameters are not isolated during charging and discharging but are interconnected; a change in one parameter often triggers a chain reaction in others. For example, increased temperature may lead to increased internal resistance, which in turn affects current stability. This complex interaction among multiple parameters makes finding a balance point in a dynamic environment exceptionally difficult. In specific business scenarios, such as during fast charging of electric vehicles, a rapid rise in temperature can lead to uneven current distribution, resulting in localized overheating, and in severe cases, even damaging the battery structure and posing safety hazards.
[0005] Therefore, how to capture and coordinate the dynamic changes among multiple parameters in real time during the charging and discharging process to ensure the stability and safety of the battery under various operating conditions has become a key issue in optimizing the dynamic charging and discharging strategy of lithium batteries. Summary of the Invention
[0006] The purpose of this invention is to propose a method for optimizing the dynamic charging and discharging strategy of lithium batteries based on multi-dimensional parameter fusion, so as to solve the problems existing in the prior art.
[0007] To achieve the above objectives, the present invention provides the following solution:
[0008] An optimization method for dynamic charging and discharging strategies of lithium batteries based on multi-dimensional parameter fusion, comprising:
[0009] Collect charging and discharging data of lithium batteries; wherein, the charging and discharging data includes: voltage parameter values, current parameter values, and temperature parameter values;
[0010] The charge and discharge data are filtered and fused with the internal resistance change rate and current stability index to obtain a fused parameter set;
[0011] Based on the fused parameter set, the particle swarm optimization algorithm is used to search for the optimal current adjustment scheme under dynamic coordination.
[0012] Based on the optimal current adjustment scheme, a dynamic and coordinated feedback loop is determined, and the state prediction of the fused parameter set is updated using the Kalman filter algorithm to obtain the updated parameter set.
[0013] By updating the parameter set, if the safety hazard risk is lower than the preset threshold, the final coordination command for the dynamic charging and discharging strategy of the lithium battery is output.
[0014] Optionally, based on the optimal current adjustment scheme, determining the dynamically changing and coordinated feedback loop includes:
[0015] Based on the optimal current adjustment scheme, obtain the optimized current value;
[0016] Based on the optimized current value, the trend of the current internal resistance change rate is obtained. If the trend shows that the local overheating phenomenon is aggravated, the control boundary of the temperature parameter value is adjusted to obtain the adjustment boundary set.
[0017] Based on adjusting the boundary set, a feedback loop for coordinating dynamic changes is determined.
[0018] Optionally, filtering the charge / discharge data includes:
[0019] The Kalman filter algorithm is used to suppress noise in voltage, current and temperature values, update the state estimates of each parameter, and generate a preliminary set of filtering parameters.
[0020] Based on the initial set of filtering parameters, the ratio between voltage and current values is calculated. If the ratio exceeds the preset threshold range, the abnormal data points are smoothed to obtain a corrected set of filtering parameters.
[0021] By modifying the filter parameter set, the correlation between the historical fluctuation range of the current value and the temperature value is analyzed. If the fluctuation range and the temperature value show a positive correlation trend, the relevant data segments are marked to determine the key influence range.
[0022] For the data within the key influence range, the characteristic value of the rate of change of internal resistance is extracted, and combined with the stability index of the current value, the dynamic relationship between the two is judged to obtain the correlation assessment result.
[0023] Based on the correlation assessment results, a comprehensive parameter set reflecting the dynamic relationship between the rate of change of internal resistance and current stability is generated.
[0024] Optionally, by integrating the internal resistance change rate and current stability indices, a fused parameter set is obtained, including:
[0025] Based on the filtered charge and discharge data, if the temperature parameter value exceeds the preset threshold, the internal resistance change rate and current stability index are integrated to obtain the integrated parameter set.
[0026] Optionally, the particle swarm optimization algorithm can be used to search for the optimal current adjustment scheme under dynamic coordination, including:
[0027] The current values of voltage and temperature parameters are extracted from the fused parameter set. A preliminary comparison is performed on the parameter values. If the voltage or temperature parameter exceeds the preset threshold range, the subsequent processing flow is triggered to obtain a preliminary anomaly judgment result.
[0028] Based on the preliminary anomaly assessment results, the corresponding dynamic change data is extracted from the data environment, and the trend of dynamic change is evaluated using a pre-established analysis model to determine whether the change is in a stable state.
[0029] Based on the dynamic and changing steady-state evaluation results, reference data for current adjustment is obtained. By filtering and processing the reference data, it is determined whether the triggering conditions of the adjustment strategy are met, and a preliminary scheme for current adjustment is obtained.
[0030] Based on the preliminary current adjustment plan, and combined with real-time data feedback on voltage and temperature parameters, the particle swarm optimization algorithm is used to search for the optimal strategy and determine the specific values of the optimization parameters.
[0031] For specific values of the optimization parameters, generate corresponding adjustment strategy data and obtain feedback on the execution status of the adjustment strategy;
[0032] Based on the feedback from the execution status of the adjustment strategy, the changes in voltage and temperature parameters in the real-time data are continuously monitored. If the parameter values exceed the preset threshold again, the dynamic change evaluation process is re-triggered to obtain the updated optimized parameters.
[0033] Optionally, the control boundaries of the temperature parameter values are adjusted to obtain the adjustment boundary set, which includes:
[0034] From the optimized current value, relevant records of the internal resistance change rate are obtained. The internal resistance change rate is compared time-by-time using a pre-established analysis tool to determine whether the internal resistance change rate shows a continuous upward trend, and to obtain a preliminary evaluation result of the change trend value.
[0035] Based on the preliminary assessment results of the trend value, if the trend value exceeds the preset threshold range, relevant data of local overheating phenomena are extracted, and the distribution of local overheating phenomena is analyzed by data comparison to determine whether local overheating phenomena are concentrated in a specific area, thus obtaining the determination result of overheating distribution.
[0036] Based on the determination results of local overheating, if the overheating distribution is concentrated in a specific area, real-time data of temperature parameter values are obtained, and anomaly detection of temperature parameter values is performed using data filtering methods to determine whether the temperature parameter values are close to the upper limit of the control boundary set, thereby obtaining the critical state assessment of the temperature parameters.
[0037] For the critical state assessment of temperature parameters, if the temperature parameter value is close to the upper limit of the control boundary set, the control boundary set is dynamically updated according to the preset adjustment rules to generate the specific range of the adjustment boundary set.
[0038] Optionally, based on adjusting the boundary set, determining the feedback loop for coordinating dynamic changes includes:
[0039] Real-time data of dynamic changes are obtained from the adjusted boundary set. By comparing the distribution characteristics of the dynamic changes, it is determined whether the dynamic changes exceed the threshold range, and the evaluation result of the dynamic changes is obtained.
[0040] For the evaluation results of dynamic changes, if the dynamic changes exceed the threshold range, the relevant records of the feedback loop are extracted, and data filtering methods are used to detect outliers in the feedback loop and determine the stable state of the feedback loop.
[0041] Optionally, the state prediction of the fused parameter set is updated using the Kalman filter algorithm. The updated parameter set includes:
[0042] Based on the steady state of the feedback loop, the input data of the filtering algorithm is obtained, and the input data of the filtering algorithm is processed by the Kalman filtering algorithm to obtain the output set of the filtering algorithm.
[0043] The current values of the fusion parameters are extracted from the output set of the filtering algorithm. A comparison tool is used to analyze the degree of deviation of the fusion parameters, determine whether the fusion parameters meet the prediction requirements, and obtain the verification results of the fusion parameters.
[0044] Based on the verification results of the fusion parameters, if the fusion parameters meet the prediction requirements, the associated data of the state prediction is updated, and the associated data of the state prediction is integrated through the data matching method to determine the optimized version of the state prediction.
[0045] Based on the optimized version of the state prediction, the initial framework of the update parameters is obtained, and the initial framework of the update parameters is processed using a correction tool to obtain the final form of the update parameters.
[0046] The adaptive data for the charging and discharging scenarios are extracted from the final form of the updated parameters. The adaptive data for the charging and discharging scenarios is adjusted according to preset rules to determine the correction scheme for the charging and discharging scenarios.
[0047] The beneficial effects of this invention are as follows:
[0048] This invention acquires an initial parameter set through sensors, uses a Kalman filter algorithm for noise filtering and state estimation to obtain a filtered parameter set, and then fuses the internal resistance change rate and current stability index to form a fused parameter set after determining that the temperature exceeds a threshold. A particle swarm optimization algorithm is then used to search for the optimal current adjustment scheme to obtain the optimized current value, and the temperature control boundary is adjusted according to the changing trend to obtain an adjustment boundary set. Based on this, a feedback loop is determined and the parameter set is updated. Finally, a coordination command set is output to guide the real-time coordination mechanism. This invention can achieve adaptive correction for lithium battery charging and discharging scenarios, effectively reduce safety risks, ensure battery stability, and provide a reliable battery management solution for improving charging efficiency and safety. Attached Figure Description
[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0050] Figure 1 This is a schematic diagram of a method for optimizing the dynamic charging and discharging strategy of lithium batteries based on multi-dimensional parameter fusion, according to an embodiment of the present invention. Detailed Implementation
[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and 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.
[0052] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0053] like Figure 1 As shown, this embodiment proposes a method for optimizing the dynamic charging and discharging strategy of lithium batteries based on multi-dimensional parameter fusion, including:
[0054] Collect charging and discharging data of lithium batteries; wherein, the charging and discharging data includes: voltage parameter values, current parameter values, and temperature parameter values;
[0055] The charge and discharge data are filtered and fused with the internal resistance change rate and current stability index to obtain a fused parameter set;
[0056] Based on the fused parameter set, the particle swarm optimization algorithm is used to search for the optimal current adjustment scheme under dynamic coordination.
[0057] Based on the optimal current adjustment scheme, a dynamic and coordinated feedback loop is determined, and the state prediction of the fused parameter set is updated using the Kalman filter algorithm to obtain the updated parameter set.
[0058] By updating the parameter set, if the safety hazard risk is lower than the preset threshold, the final coordination command for the dynamic charging and discharging strategy of the lithium battery is output.
[0059] Furthermore, based on the optimal current adjustment scheme, the feedback loop for dynamic change coordination is determined to include:
[0060] Based on the optimal current adjustment scheme, obtain the optimized current value;
[0061] Based on the optimized current value, the trend of the current internal resistance change rate is obtained. If the trend shows that the local overheating phenomenon is aggravated, the control boundary of the temperature parameter value is adjusted to obtain the adjustment boundary set.
[0062] Based on adjusting the boundary set, a feedback loop for coordinating dynamic changes is determined.
[0063] Specifically, in this embodiment, the collection of lithium battery charging and discharging data involves: real-time collection of voltage, current, and temperature parameters under charging and discharging scenarios using sensor devices, recording the parameter data at each time point, and obtaining the raw dataset.
[0064] Furthermore, filtering the charge / discharge data includes:
[0065] The Kalman filter algorithm is used to suppress noise in voltage, current and temperature values, update the state estimates of each parameter, and generate a preliminary set of filtering parameters.
[0066] Based on the initial set of filtering parameters, the ratio between voltage and current values is calculated. If the ratio exceeds the preset threshold range, the abnormal data points are smoothed to obtain a corrected set of filtering parameters.
[0067] By modifying the filter parameter set, the correlation between the historical fluctuation range of the current value and the temperature value is analyzed. If the fluctuation range and the temperature value show a positive correlation trend, the relevant data segments are marked to determine the key influence range.
[0068] For the data within the key influence range, the characteristic value of the rate of change of internal resistance is extracted, and combined with the stability index of the current value, the dynamic relationship between the two is judged to obtain the correlation assessment result.
[0069] Based on the correlation assessment results, a comprehensive parameter set reflecting the dynamic relationship between the rate of change of internal resistance and current stability is generated.
[0070] Specifically, in this embodiment, a Kalman filter algorithm is used to filter noise and estimate the state of voltage, current, and temperature parameters for the initial parameter set, and a dynamic correlation of the filtered parameter set reflecting the rate of change of internal resistance and the stability of the current is obtained. For example, the implementation method is as follows: First, assume the initial parameter set is a voltage value of 3.7V, a current value of 2.5A, and a temperature value of 25°C. These data are collected by sensors but contain noise and need to be filtered. The Kalman filter algorithm is selected for this purpose, and its core is to optimize the estimated value through state prediction and measurement update iteration. Taking voltage as an example, the initial state vector is set as x0 = [3.7 0]T, representing the voltage value and rate of change, the state transition matrix F = [[1 0.1] [0 1]], the time step is 0.1 seconds, the process noise covariance Q = [[0.01 0] [0 0.001]], and the measurement noise covariance R = 0.05. In the prediction phase, the state prediction is x_pred = F * x0, and the covariance prediction is P_pred = F * P0 * F^T + Q, where P0 is the initial covariance matrix [[0.1 0] [0 0.1]]. In the measurement update phase, assuming the actual measured voltage is 3.72V, the Kalman gain K = P_pred * H^T * (H * P_pred * H^T + R)^(-1), where H =
[10] is the measurement matrix, the state is updated x_est = x_pred + K * (3.72 H * x_pred), and the covariance is updated P_est = ( K * H) * P_pred, finally obtaining a filtered voltage value of approximately 3.71V. Similarly, the current and temperature are processed by the same algorithm, resulting in a filtered current value of 2.48A and a temperature value of 25.2°C. Next, the dynamic correlation between the rate of change of internal resistance and current stability was analyzed. The rate of change of internal resistance, calculated using filtered voltage and current, was (3.71 / 2.48 + 3.7 / 2.5) / 0.1 ≈ 0.014 ohms / second. Current stability was calculated using the variance of the filtered current values. Assuming the variance of the current values over 10 consecutive time steps is 0.002, it indicates relatively stable current. Finally, the filtered parameter set was set to voltage 3.71V, current 2.48A, and temperature 25.2°C. A dynamic correlation model was established using the rate of change of internal resistance and current variance to provide data support for subsequent battery health status assessment.
[0071] Furthermore, by integrating the internal resistance change rate and current stability indices, a fusion parameter set is obtained, including:
[0072] Based on the filtered charge and discharge data, if the temperature parameter value exceeds the preset threshold, the internal resistance change rate and current stability index are integrated to obtain the integrated parameter set.
[0073] Specifically, in this embodiment, assuming the currently collected temperature is 85 degrees Celsius and the preset threshold is 80 degrees Celsius, the comparison algorithm determines that 85 is greater than 80, triggering the subsequent fusion process. The analysis logic is that the temperature exceeding the standard may cause local overheating risk, therefore, it is necessary to further integrate other indicators for comprehensive evaluation. Next, the internal resistance change rate data is automatically extracted. Assuming the current internal resistance change rate is 5.2% and the historical average is 3.0%, the change rate deviation is calculated as (5.2-3.0) / 3.0=0.733, i.e., the deviation rate is 73.3%, indicating an abnormal increase in internal resistance, which may be related to overheating. This deviation rate is considered as one of the risk factors. At the same time, the current stability index is collected. Assuming the current fluctuation value is 2.5 amperes and the standard fluctuation threshold is 1.0 amperes, the fluctuation ratio is calculated as 2.5 / 1.0=2.5, indicating poor current stability, which may exacerbate the overheating risk. Logically, this is related to temperature and internal resistance, because unstable current increases the probability of heating. Finally, the parameters are integrated through a weighted fusion algorithm.
[0074] Furthermore, a particle swarm optimization algorithm is used to search for the optimal current adjustment scheme under dynamic coordination, including:
[0075] The current values of voltage and temperature parameters are extracted from the fused parameter set. A preliminary comparison is performed on the parameter values. If the voltage or temperature parameter exceeds the preset threshold range, the subsequent processing flow is triggered to obtain a preliminary anomaly judgment result.
[0076] Based on the preliminary anomaly assessment results, the corresponding dynamic change data is extracted from the data environment, and the trend of dynamic change is evaluated using a pre-established analysis model to determine whether the change is in a stable state.
[0077] Based on the dynamic and changing steady-state evaluation results, reference data for current adjustment is obtained. By filtering and processing the reference data, it is determined whether the triggering conditions of the adjustment strategy are met, and a preliminary scheme for current adjustment is obtained.
[0078] Based on the preliminary current adjustment plan, and combined with real-time data feedback on voltage and temperature parameters, the particle swarm optimization algorithm is used to search for the optimal strategy and determine the specific values of the optimization parameters.
[0079] For specific values of the optimization parameters, generate corresponding adjustment strategy data and obtain feedback on the execution status of the adjustment strategy;
[0080] Based on the feedback from the execution status of the adjustment strategy, the changes in voltage and temperature parameters in the real-time data are continuously monitored. If the parameter values exceed the preset threshold again, the dynamic change evaluation process is re-triggered to obtain the updated optimized parameters.
[0081] Specifically, in this embodiment, assuming the current voltage is 24.5 volts and the temperature is 78 degrees Celsius, these two parameters are compared with historical safe ranges. The safe voltage range is 22.0 to 26.0 volts, and the safe temperature range is 60 to 80 degrees Celsius. The algorithm determines that the current values are within the range, but the temperature is close to the upper limit, requiring further optimization of the current to avoid potential risks. Next, the particle swarm optimization algorithm is called, initializing the particle swarm size to 50 particles, with each particle representing a current adjustment scheme. The initial current value range is set to 10.0 to 20.0 amperes. The objective function is defined as minimizing the combined deviation between temperature and voltage, with the formula: target value = 0.6 * (temperature - 70) / 70 + 0.4 * (voltage - 24) / 24. The calculation is iterated 10 times, updating the particle position and velocity each time. Assuming that after the 10th iteration, the current value corresponding to the optimal particle is 15.3 amperes, the target value drops to 0.12, indicating that this scheme is relatively ideal. Subsequently, based on real-time feedback, parameters are adjusted, with 15.3 amps selected as the optimized current value. The analysis logic aims to reduce the upward trend of temperature while maintaining voltage stability and minimizing safety hazards. If subsequent monitoring detects that the temperature continues to approach 80 degrees Celsius, a new round of optimization will be automatically triggered. To ensure the integrity of the logic, the optimized current value is matched with the equipment load requirements. Assuming the load requirement is 14.8 to 16.0 amps, 15.3 amps meets the requirements. This optimization result is automatically recorded, and control commands are updated, forming a closed-loop processing chain from parameter acquisition to current adjustment.
[0082] Furthermore, by adjusting the control boundaries of the temperature parameter values, the set of adjustment boundaries is obtained, including:
[0083] From the optimized current value, relevant records of the internal resistance change rate are obtained. The internal resistance change rate is compared time-by-time using a pre-established analysis tool to determine whether the internal resistance change rate shows a continuous upward trend, and to obtain a preliminary evaluation result of the change trend value.
[0084] Based on the preliminary assessment results of the trend value, if the trend value exceeds the preset threshold range, relevant data of local overheating phenomena are extracted, and the distribution of local overheating phenomena is analyzed by data comparison to determine whether local overheating phenomena are concentrated in a specific area, thus obtaining the determination result of overheating distribution.
[0085] Based on the determination results of local overheating, if the overheating distribution is concentrated in a specific area, real-time data of temperature parameter values are obtained, and anomaly detection of temperature parameter values is performed using data filtering methods to determine whether the temperature parameter values are close to the upper limit of the control boundary set, thereby obtaining the critical state assessment of the temperature parameters.
[0086] For the critical state assessment of temperature parameters, if the temperature parameter value is close to the upper limit of the control boundary set, the control boundary set is dynamically updated according to the preset adjustment rules to generate the specific range of the adjustment boundary set.
[0087] Specifically, in this embodiment, starting from the optimized current value, the trend of the current internal resistance change rate is automatically calculated. Assuming the optimized current value is 16.2 amperes, this embodiment collects internal resistance data over the past 30 minutes using a built-in sensor, showing that the internal resistance value increased from 0.5 ohms to 0.55 ohms, with a calculated change rate of 10%. Combining this with a historical data analysis model, a change rate threshold of 8% is set. The current value exceeds the threshold, and this embodiment determines that there is a possibility of increased local overheating. Next, a trend prediction algorithm is invoked. Based on the time series data of the internal resistance change rate, a linear regression method is used to fit the internal resistance change curve for the next 10 minutes. The prediction results show that the internal resistance may further increase to 0.58 ohms, indicating that the overheating risk persists and relevant parameters need to be adjusted. Subsequently, the control boundaries for the temperature parameter values are dynamically adjusted. The initial temperature boundary is 65 to 85 degrees Celsius. Combining the rate of change of internal resistance and the prediction results, the new upper limit is calculated as 85 - 10 * 0.5 = 80 degrees Celsius using a preset boundary adjustment formula: new upper limit = current upper limit - rate of change * 5. The lower limit remains unchanged, forming an adjustment boundary set of 65 to 80 degrees Celsius. Finally, in this embodiment, the adjustment boundary set is applied to the current stability control module. By monitoring the matching degree between the temperature data and the new boundary in real time, a control signal is automatically generated. If the temperature is close to 80 degrees Celsius, stability is maintained by reducing the current output power. Assuming the current temperature is 79 degrees Celsius, the calculated power reduction ratio is 5%, thereby ensuring that the equipment operates within a safe range. At the same time, the adjustment process data is recorded to the database, forming a complete logical chain from internal resistance change to boundary adjustment.
[0088] Furthermore, based on adjusting the boundary set, the feedback loop for coordinating dynamic changes is determined to include:
[0089] Real-time data of dynamic changes are obtained from the adjusted boundary set. By comparing the distribution characteristics of the dynamic changes, it is determined whether the dynamic changes exceed the threshold range, and the evaluation result of the dynamic changes is obtained.
[0090] For the evaluation results of dynamic changes, if the dynamic changes exceed the threshold range, the relevant records of the feedback loop are extracted, and data filtering methods are used to detect outliers in the feedback loop and determine the stable state of the feedback loop.
[0091] Furthermore, the state prediction of the fused parameter set is updated using the Kalman filter algorithm. The updated parameter set includes:
[0092] Based on the steady state of the feedback loop, the input data of the filtering algorithm is obtained, and the input data of the filtering algorithm is processed by the Kalman filtering algorithm to obtain the output set of the filtering algorithm.
[0093] The current values of the fusion parameters are extracted from the output set of the filtering algorithm. A comparison tool is used to analyze the degree of deviation of the fusion parameters, determine whether the fusion parameters meet the prediction requirements, and obtain the verification results of the fusion parameters.
[0094] Based on the verification results of the fusion parameters, if the fusion parameters meet the prediction requirements, the associated data of the state prediction is updated, and the associated data of the state prediction is integrated through the data matching method to determine the optimized version of the state prediction.
[0095] Based on the optimized version of the state prediction, the initial framework of the update parameters is obtained, and the initial framework of the update parameters is processed using a correction tool to obtain the final form of the update parameters.
[0096] The adaptive data for the charging and discharging scenarios are extracted from the final form of the updated parameters. The adaptive data for the charging and discharging scenarios is adjusted according to preset rules to determine the correction scheme for the charging and discharging scenarios.
[0097] Specifically, in this embodiment, a feedback loop is first constructed by dynamically adjusting the boundary conditions. Assuming the initial boundary set is a voltage range of 3.2V to 4.2V and a current limit of 5A to 10A, the boundaries are dynamically adjusted based on real-time collected charging data (e.g., voltage 3.8V, current 7A). If the voltage approaches the upper limit of 4.2V, the charging current is reduced to 6.5A through the feedback loop to avoid overvoltage. The adjusted boundary data is recorded as new constraints, forming a closed-loop control. Next, the Kalman filter algorithm is used to predict and update the fused parameter set. Assuming the initial state parameter set includes a charging efficiency of 0.85 and a temperature coefficient of 0.02, the noise covariance matrix Q is set to 0.01, and the measurement noise R is 0.05, the gain K = 0.17 is obtained using the Kalman gain calculation formula K = P / (P + R). The state prediction value is then updated, correcting the charging efficiency to 0.87, and the temperature coefficient is adjusted to 0.021 based on actual measured values (e.g., a temperature increase of 2.3 degrees), forming the updated parameter set. For adaptive modifications to fast charging scenarios, parameters are further optimized based on charging rate requirements (such as achieving 80% charge within 30 minutes). If the battery internal resistance is detected to rise to 0.05 ohms, the charging current is reduced to 6A through an algorithm to reduce heat generation, while the charging efficiency parameter is adjusted to 0.86 to ensure the safety of fast charging.
[0098] This embodiment acquires an initial parameter set through sensors, uses a Kalman filter algorithm for noise filtering and state estimation to obtain a filtered parameter set, and then fuses the internal resistance change rate and current stability index to form a fused parameter set after determining that the temperature exceeds a threshold. A particle swarm optimization algorithm is used to search for the optimal current adjustment scheme to obtain the optimized current value, and the temperature control boundary is adjusted according to the changing trend to obtain an adjustment boundary set. Based on this, a feedback loop is determined and the parameter set is updated. Finally, a coordination command set is output to guide the real-time coordination mechanism. This invention can achieve adaptive correction for lithium battery charging and discharging scenarios, effectively reduce safety risks, ensure battery stability, and provide a reliable battery management solution for improving charging efficiency and safety.
[0099] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.
Claims
1. A method for optimizing dynamic charging and discharging strategies of lithium batteries based on multi-dimensional parameter fusion, characterized in that, include: Collect charging and discharging data of lithium batteries; wherein, the charging and discharging data includes: voltage parameter values, current parameter values, and temperature parameter values; The charge and discharge data are filtered and fused with the internal resistance change rate and current stability index to obtain a fused parameter set; Based on the fused parameter set, the particle swarm optimization algorithm is used to search for the optimal current adjustment scheme under dynamic coordination. Based on the optimal current adjustment scheme, a dynamic and coordinated feedback loop is determined, and the state prediction of the fused parameter set is updated using the Kalman filter algorithm to obtain the updated parameter set. By updating the parameter set, if the safety hazard risk is lower than the preset threshold, the final coordination command for the dynamic charging and discharging strategy of the lithium battery is output.
2. The method for optimizing dynamic charging and discharging strategies of lithium batteries based on multi-dimensional parameter fusion according to claim 1, characterized in that, Based on the optimal current adjustment scheme, the feedback loop for dynamic coordination is determined as follows: Based on the optimal current adjustment scheme, obtain the optimized current value; Based on the optimized current value, the trend of the current internal resistance change rate is obtained. If the trend shows that the local overheating phenomenon is aggravated, the control boundary of the temperature parameter value is adjusted to obtain the adjustment boundary set. Based on adjusting the boundary set, a feedback loop for coordinating dynamic changes is determined.
3. The method for optimizing dynamic charging and discharging strategies of lithium batteries based on multi-dimensional parameter fusion according to claim 1, characterized in that, Filtering the charge and discharge data includes: The Kalman filter algorithm is used to suppress noise in voltage, current and temperature values, update the state estimates of each parameter, and generate a preliminary set of filtering parameters. Based on the initial set of filtering parameters, the ratio between voltage and current values is calculated. If the ratio exceeds the preset threshold range, the abnormal data points are smoothed to obtain a corrected set of filtering parameters. By modifying the filter parameter set, the correlation between the historical fluctuation range of the current value and the temperature value is analyzed. If the fluctuation range and the temperature value show a positive correlation trend, the relevant data segments are marked to determine the key influence range. For the data within the key influence range, the characteristic value of the rate of change of internal resistance is extracted, and combined with the stability index of the current value, the dynamic relationship between the two is judged to obtain the correlation assessment result. Based on the correlation assessment results, a comprehensive parameter set reflecting the dynamic relationship between the rate of change of internal resistance and current stability is generated.
4. The method for optimizing dynamic charging and discharging strategies of lithium batteries based on multi-dimensional parameter fusion according to claim 1, characterized in that, By integrating the internal resistance change rate and current stability indices, a fusion parameter set is obtained, including: Based on the filtered charge and discharge data, if the temperature parameter value exceeds the preset threshold, the internal resistance change rate and current stability index are integrated to obtain the integrated parameter set.
5. The method for optimizing dynamic charging and discharging strategies of lithium batteries based on multi-dimensional parameter fusion according to claim 1, characterized in that, The particle swarm optimization algorithm is used to search for the optimal current adjustment scheme under dynamic coordination, including: The current values of voltage and temperature parameters are extracted from the fused parameter set. A preliminary comparison is performed on the parameter values. If the voltage or temperature parameter exceeds the preset threshold range, the subsequent processing flow is triggered to obtain a preliminary anomaly judgment result. Based on the preliminary anomaly assessment results, the corresponding dynamic change data is extracted from the data environment, and the trend of dynamic change is evaluated using a pre-established analysis model to determine whether the change is in a stable state. Based on the dynamic and changing steady-state evaluation results, reference data for current adjustment is obtained. By filtering and processing the reference data, it is determined whether the triggering conditions of the adjustment strategy are met, and a preliminary scheme for current adjustment is obtained. Based on the preliminary current adjustment plan, and combined with real-time data feedback on voltage and temperature parameters, the particle swarm optimization algorithm is used to search for the optimal strategy and determine the specific values of the optimization parameters. For specific values of the optimization parameters, generate corresponding adjustment strategy data and obtain feedback on the execution status of the adjustment strategy; Based on the feedback from the execution status of the adjustment strategy, the changes in voltage and temperature parameters in the real-time data are continuously monitored. If the parameter values exceed the preset threshold again, the dynamic change evaluation process is re-triggered to obtain the updated optimized parameters.
6. The method for optimizing dynamic charging and discharging strategies of lithium batteries based on multi-dimensional parameter fusion according to claim 2, characterized in that, Adjusting the control boundaries of the temperature parameter values yields the following adjustment boundary set: From the optimized current value, relevant records of the internal resistance change rate are obtained. The internal resistance change rate is compared time-by-time using a pre-established analysis tool to determine whether the internal resistance change rate shows a continuous upward trend, and to obtain a preliminary evaluation result of the change trend value. Based on the preliminary assessment results of the trend value, if the trend value exceeds the preset threshold range, relevant data of local overheating phenomena are extracted, and the distribution of local overheating phenomena is analyzed by data comparison to determine whether local overheating phenomena are concentrated in a specific area, thus obtaining the determination result of overheating distribution. Based on the determination results of local overheating, if the overheating distribution is concentrated in a specific area, real-time data of temperature parameter values are obtained, and anomaly detection of temperature parameter values is performed using data filtering methods to determine whether the temperature parameter values are close to the upper limit of the control boundary set, thereby obtaining the critical state assessment of the temperature parameters. For the critical state assessment of temperature parameters, if the temperature parameter value is close to the upper limit of the control boundary set, the control boundary set is dynamically updated according to the preset adjustment rules to generate the specific range of the adjustment boundary set.
7. The method for optimizing dynamic charging and discharging strategies of lithium batteries based on multi-dimensional parameter fusion according to claim 2, characterized in that, Based on adjusting the boundary set, the feedback loop for coordinating dynamic changes is determined to include: Real-time data of dynamic changes are obtained from the adjusted boundary set. By comparing the distribution characteristics of the dynamic changes, it is determined whether the dynamic changes exceed the threshold range, and the evaluation result of the dynamic changes is obtained. For the evaluation results of dynamic changes, if the dynamic changes exceed the threshold range, the relevant records of the feedback loop are extracted, and data filtering methods are used to detect outliers in the feedback loop and determine the stable state of the feedback loop.
8. The method for optimizing dynamic charging and discharging strategies of lithium batteries based on multi-dimensional parameter fusion according to claim 7, characterized in that, The state prediction of the fused parameter set is updated using the Kalman filter algorithm. The updated parameter set includes: Based on the steady state of the feedback loop, the input data of the filtering algorithm is obtained, and the input data of the filtering algorithm is processed by the Kalman filtering algorithm to obtain the output set of the filtering algorithm. The current values of the fusion parameters are extracted from the output set of the filtering algorithm. A comparison tool is used to analyze the degree of deviation of the fusion parameters, determine whether the fusion parameters meet the prediction requirements, and obtain the verification results of the fusion parameters. Based on the verification results of the fusion parameters, if the fusion parameters meet the prediction requirements, the associated data of the state prediction is updated, and the associated data of the state prediction is integrated through the data matching method to determine the optimized version of the state prediction. Based on the optimized version of the state prediction, the initial framework of the update parameters is obtained, and the initial framework of the update parameters is processed using a correction tool to obtain the final form of the update parameters. The adaptive data for the charging and discharging scenarios are extracted from the final form of the updated parameters. The adaptive data for the charging and discharging scenarios is adjusted according to preset rules to determine the correction scheme for the charging and discharging scenarios.