Charging and discharging control system and method based on dynamic mathematical model

Abstract

The invention discloses a charge and discharge control system and method based on a dynamic mathematical model, and relates to the technical field of lithium batteries, the charge and discharge control system comprises a model modeling module, a prediction and strategy module and an execution and guardrail module, the model modeling module is used for collecting voltage, current, temperature and time information, estimating the state of charge, the state of health and thermal-electric parameters on line, and calculating the state of charge, the state of health and the thermal-electric parameters; the control module is used for generating an electric, thermal and aging dynamic model, the prediction and strategy module is used for generating a rolling optimization track with charge and discharge power and balance cooperation based on the model and a safety guardrail and outputting a target current, a balance instruction and a temperature management strategy, and the execution and guardrail module is used for driving a charge and discharge channel and a balance branch according to the target instruction. The method comprises the following steps of: S1, collecting monomer and total voltage, current, temperature and timing information at a starting stage, and performing grading over-charging, over-discharging, over-current, short circuit and temperature protection and overtime alarm, recording a state and feeding back the state to a model; the method comprises the following steps of: S1, collecting monomer and total voltage, current, temperature and timing information at a starting stage;

Description

Technical Field

[0001] This invention relates to the field of lithium battery technology, specifically to a charging and discharging control system and method based on a dynamic mathematical model. Background Technology

[0002] The lithium battery management system is responsible for detecting and protecting voltage, current, temperature, and capacity, and performing functions such as overcharge, over-discharge, overcurrent, short circuit and temperature protection, equalization, and state memory during charging and discharging. Existing products mostly use fixed thresholds and simple logic to trigger protection and recovery (such as graded overcharge, graded overcurrent, rapid short circuit cutoff, charging and discharging temperature range limitation, equalization voltage difference threshold, long-term charging timeout alarm, state memory before and after sleep mode, and power indication, etc.), which can cover most basic scenarios. However, under conditions such as rapid temperature changes, uneven aging, large load fluctuations, fast charging, and low-temperature start-up and discharge, fixed thresholds and static strategies are difficult to achieve a comprehensive optimization of safety boundaries, lifespan, and efficiency.

[0003] In existing technologies, fixed voltage, current, and temperature thresholds, along with fixed delays, cannot be dynamically adjusted according to environmental conditions and aging processes, easily leading to conservative shutdowns and boundary jitter. Voltage, temperature, internal resistance, and aging factors interact with each other, making it difficult for simple logic to quantify impending boundary exceedances. Furthermore, the lack of goal-oriented trajectory optimization, coupled with the disconnect between balancing, charging current, and heat dissipation strategies, limits the availability of fast charging and low-temperature applications. Therefore, designing a charge / discharge control system and method based on a dynamic mathematical model is essential. Summary of the Invention

[0004] The purpose of this invention is to provide a charging and discharging control system and method based on a dynamic mathematical model to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a charge-discharge control system based on a dynamic mathematical model, comprising a modeling module, a prediction and strategy module, and an execution and guardrail module. The modeling module is used to collect voltage, current, temperature, and time information, estimate the state of charge, health status, and thermo-electrical parameters online, and form a dynamic model of electrical, thermal, and aging parameters. The prediction and strategy module generates a rolling optimization trajectory for coordinated charging and discharging power and equalization based on the model and safety guardrail, and outputs target current, equalization commands, and temperature management strategies. The execution and guardrail module drives the charging and discharging channels and equalization branches according to the target commands, executes graded overcharge, over-discharge, overcurrent, short circuit, and temperature protection, and overtime alarms, records the status, and feeds back to the model.

[0006] According to the above technical solution, the modeling module includes a parameter acquisition module, a state estimation module, and a model update module. The parameter acquisition module is connected to the cell assembly and temperature sensor via a signal to acquire information on individual cell and total voltage, current, temperature, and timing. The state estimation module is connected to the parameter acquisition module via a signal to estimate SOC, SOH, equivalent internal resistance, and thermal parameters. The model update module is connected to the state estimation module via a signal to perform rolling updates of model parameters based on historical records and alarm counts. The prediction and strategy module includes a guardrail generation module, a trajectory planning module, and a power allocation module. The guardrail generation module is connected to the model update module via a signal and is used to generate soft guardrails and hysteresis windows outside a fixed threshold. The trajectory planning module is connected to the guardrail generation module and the state estimation module via a signal and is used to plan the charging and discharging current and the equalization current in a rolling manner within constraints. The power allocation module is connected to the trajectory planning module via a control and is used to convert the trajectory into real-time instructions for charging, discharging, and equalization. The execution and guardrail module includes a power execution module, a protection control module, and an indication and memory module. The power execution module is electrically connected to the charging and discharging channel and the equalization branch, and is used to output according to instructions and sample and read back in real time. The protection control module is electrically connected to the measurement front end and the power switch, and is used to execute graded overcharge, over-discharge, overcurrent, short circuit and temperature protection and recovery strategies. The indication and memory module is connected to the protection control module and the model update module via signals, and is used for power indication, alarm display, memory of state before and after sleep mode and count of times.

[0007] A charge-discharge control method based on a dynamic mathematical model, comprising the following steps: S1. During the startup phase, collect individual and total voltage, current, temperature and timing information, combine historical samples to obtain initial estimates of SOC, SOH and equivalent internal resistance and thermal parameters, generate an integrated dynamic model of electrical, thermal and aging. During operation, use readback data and event records, including overcurrent, overtemperature, timeout, and equalization triggers, to perform rolling correction of model parameters, output the current health status and prediction boundary as input for subsequent guardrail and trajectory planning. S2. In addition to the overcharge, over-discharge, overcurrent, short circuit and temperature hard thresholds specified in the specifications, an early warning zone and hysteresis window are automatically generated near the boundary to form the allowable power range that can be executed under the current operating conditions. The range is then shrunk and widened according to the scenario based on the ambient temperature, aging degree and differential pressure, and linked with the timing and recovery logic to reduce boundary back-and-forth triggering and false recovery. S3. Generate the target current trajectory for charging and discharging within the allowable range, and use balancing as a cooperative variable to make the voltage difference between cells tend to converge without affecting the main power path. Check the temperature, voltage difference, timeout and recovery conditions of each candidate trajectory, eliminate the schemes with risks of overheating and overcooling, voltage difference exceeding the limit and charging timeout, and output the executable trajectory and balancing command. S4. Drive the power and equalization branch according to the target current and equalization command, sample and read back in real time, and adaptively switch between small-step fine-tuning and trying to cross the level according to the improvement rate and group diversity. After crossing the level, recalculate the subordinate parameters and location labels synchronously. If overcurrent, short circuit, overtemperature, timeout and voltage difference abnormality are encountered, immediately cut off and reduce the rate according to the graded strategy, and record the event for subsequent model fine-tuning and guardrail correction. S5. When the preset time is reached and there are insufficient improvements in multiple consecutive cycles, and the hibernation condition is triggered, the current cycle ends. The strategy that passes all hard boundary checks and has the best comprehensive score is selected as the final result. The equalization record, power and temperature, time curve, key event statistics and energy efficiency assessment are output simultaneously to provide a basis for subsequent cycle power-on initialization and remote maintenance.

[0008] According to the above technical solution, S1 specifically refers to: S1-1. During power-on and initial operation, acquire individual and total voltage, current, temperature and timing information according to the preset sampling cycle, complete range and zero drift correction, and perform noise reduction and outlier removal; combine historical samples and factory parameters to give initial estimates of SOC, SOH, equivalent internal resistance and key thermal parameters, generate an integrated dynamic model of electrical, thermal and aging, and mark the version number, timestamp and applicable temperature zone as the baseline for subsequent guardrail generation and trajectory planning; S1-2. During closed-loop execution, continuously receive readback data and event tags, including over-temperature, over-current, timeout, and equalization trigger. Perform small-step rolling correction and temperature-segmented correction on model parameters. Track the voltage and temperature differences between cells to assess consistency and degradation trends. Set isolation markers and reduce weights for channels with drift anomalies and sudden noise increases, and synchronously publish the updated state variables and boundary predictions to subsequent stages.

[0009] According to the above technical solution, S2 specifically refers to: S2-1. Without changing the hard thresholds and recovery conditions for overcharge, over-discharge, overcurrent, short circuit, and temperature specified in the datasheet, generate warning bands and hysteresis windows near the boundaries for key variables, and clearly define the allowable power range and recovery path; align the timing and alarm rules with this range to form an executable and implementable charge / discharge power-temperature joint constraint package, and output it for trajectory planning. S2-2. Combining ambient temperature, current State of Health (SOH), consistency level, and load scenario, the allowable range is adaptively shrunk and widened according to the scenario, and the start / stop thresholds of the balancing strategy and heat dissipation strategy are linked; a buffer zone is added for conditions near the boundary and the safety margin is recorded, outputting the final available power range and strategy trigger threshold, reducing the possibility of boundary back-and-forth triggering and false recovery, when the power change value Less than the lower critical value If this occurs, the scenario is defined as a potentially low safety margin scenario, and the voltage rebound amplitude is collected. The variance is calculated, if the voltage rebound amplitude variance Greater than the set value If so, it is determined to be in an overcharge protection state.

[0010] According to the above technical solution, S3 specifically refers to: S3-1. Within the allowable range, generate the target current trajectory using a rolling time window, constrain the current change rate, and link it with equalization as a co-variable to bring the voltage difference between cells closer to the allowable range. For different scenarios, including fast charging, low-temperature start-up and discharge, and long-term float charging, provide differentiated trajectory templates and initial values ​​to ensure improved energy utilization and temperature control efficiency without compromising safety. Locate the regions corresponding to high polarization and high internal resistance zones in the cell group state distribution map and electrochemical characteristic map, and increase the equalization trigger interval in these regions. The magnitude of the increase and and The ratios are positively correlated, specifically: ,in The default trigger interval is... This is the weighting coefficient for the voltage rebound amplitude and rate; S3-2. Verify the temperature, differential pressure, timing and recovery conditions of each candidate trajectory, and eliminate schemes with risks of overheating or overcooling, differential pressure exceeding limits and charging timeout; automatically reduce the derating of schemes close to the boundary and provide backup trajectories, and finally output the executable main trajectory and equalization command, with execution priority and rollback conditions to ensure stable landing under field fluctuations and measurement disturbances.

[0011] According to the above technical solution, S4 specifically refers to: S4-1: Drive the power and equalization branch according to the target current and equalization command, compare the target and readback in real time and maintain within a small error range; when the improvement rate is good and far from the boundary, enter small-step refinement, reduce the disturbance amplitude to reduce costs and increase safety margin; when continuous improvement is insufficient and close to the upper edge, enter the attempt to cross the level, apply adsorption reward, amplify the disturbance amplitude and limit the exploration to a better direction, and immediately recalculate the subordinate parameters and location labels after the transition to maintain stability; S4-2. When overcurrent, short circuit, overtemperature, timeout and voltage difference abnormalities occur, quickly cut off and reduce the derating according to the graded strategy, and record the number of events, duration and recovery path; after recovery, make small-step corrections to the relevant thresholds and model parameters, and mark the key segments and equalization effects of this round of execution as inheritable segments, as the empirical basis for the next generation initial solution and subsequent batch initialization.

[0012] According to the above technical solution, S5 specifically refers to: S5-1. When the preset time and loop count limit are reached, and there are insufficient improvements within a certain number of consecutive time windows, and the condition for entering hibernation is met, the current optimization and execution loop is terminated. Before termination, a quick feasibility review and data integrity check are performed to ensure that the output results can be reproduced and audited. S5-2 outputs the final charge / discharge strategy, equalization execution record, temperature-time and voltage-time curves, key event statistics and energy efficiency assessment, and marks the applicable temperature range and health status range; at the same time, it writes to memory and remote logs, and provides initialization suggestions and parameter starting points for the next power-on, so as to be reused and maintained across batches and platforms.

[0013] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: The present invention proposes a dynamic mathematical model-driven charge and discharge control system: under the premise of meeting the existing multi-level protection and temperature, equalization and timing requirements of the specification, online parameter identification and state estimation are introduced to construct an integrated dynamic model of the battery cell pack's electrical-thermal-aging, generate the optimal trajectory of charge and discharge power and equalization coordination according to the model prediction, and suppress boundary jitter with multi-layer safety guardrails and double threshold hysteresis, and finally achieve closed-loop control that maximizes efficiency under safety and life constraints; Without altering the existing hard threshold system, prediction and soft guardrails are added to significantly reduce oscillations and false touches under critical conditions. The charging current and balancing current are dynamically allocated to reduce overcharging, overheating shutdowns, and deep over-discharge, thereby improving effective capacity utilization and cycle life. Threshold fine-tuning and trajectory optimization are driven by models and historical statistics, and are adaptive to cell consistency and aging status. The maintenance cost is low, and it is portable across platforms. Attached Figure Description

[0014] 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 schematic diagram of the overall modular structure of the present invention. Detailed Implementation

[0015] 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.

[0016] Please see Figure 1 This invention provides a technical solution: a charging and discharging control system and method based on a dynamic mathematical model, including a model modeling module, a prediction and strategy module, and an execution and guardrail module. The model modeling module is used to collect voltage, current, temperature and time information, estimate the state of charge, health state and thermo-electric parameters online, and form a dynamic model of electrical, thermal and aging. The prediction and strategy module generates a rolling optimization trajectory of charging and discharging power and equalization coordination based on the model and safety guardrail, and outputs the target current, equalization command and temperature management strategy. The execution and guardrail module drives the charging and discharging channel and equalization branch according to the target command, performs graded overcharge, over-discharge, overcurrent, short circuit and temperature protection and timeout alarm, records the status and feeds it back to the model. The modeling module includes a parameter acquisition module, a state estimation module, and a model update module. The parameter acquisition module is connected to the cell assembly and temperature sensor via a signal to acquire information on individual cell and total voltage, current, temperature, and timing. The state estimation module is connected to the parameter acquisition module via a signal to estimate SOC, SOH, equivalent internal resistance, and thermal parameters. The model update module is connected to the state estimation module via a signal to perform rolling updates of model parameters based on historical records and alarm counts. The prediction and strategy module includes a guardrail generation module, a trajectory planning module, and a power allocation module. The guardrail generation module is connected to the model update module via a signal to generate soft guardrails and hysteresis windows outside a fixed threshold. The trajectory planning module is connected to the guardrail generation module and the state estimation module via a signal to plan the charging and discharging current and the equalization current in a rolling manner within constraints. The power allocation module is connected to the trajectory planning module via a control to convert the trajectory into real-time instructions for charging, discharging, and equalization. The execution and guardrail module includes a power execution module, a protection control module, and an indication and memory module. The power execution module is electrically connected to the charging and discharging channel and the equalization branch, and is used to output according to the command and sample and read back in real time. The protection control module is electrically connected to the measurement front end and the power switch, and is used to execute graded overcharge, over-discharge, overcurrent, short circuit and temperature protection and recovery strategies. The indication and memory module is connected to the protection control module and the model update module via a signal, and is used for power indication, alarm display, memory of the state before and after sleep and count of times. Includes the following steps: S1. During the startup phase, collect individual and total voltage, current, temperature and timing information, combine historical samples to obtain initial estimates of SOC, SOH and equivalent internal resistance and thermal parameters, generate an integrated dynamic model of electrical, thermal and aging. During operation, use readback data and event records, including overcurrent, overtemperature, timeout, and equalization triggers, to perform rolling correction of model parameters, output the current health status and prediction boundary as input for subsequent guardrail and trajectory planning. S2. In addition to the overcharge, over-discharge, overcurrent, short circuit and temperature hard thresholds specified in the specifications, an early warning zone and hysteresis window are automatically generated near the boundary to form the allowable power range that can be executed under the current operating conditions. The range is then shrunk and widened according to the scenario based on the ambient temperature, aging degree and differential pressure, and linked with the timing and recovery logic to reduce boundary back-and-forth triggering and false recovery. S3. Generate the target current trajectory for charging and discharging within the allowable range, and use balancing as a cooperative variable to make the voltage difference between cells tend to converge without affecting the main power path. Check the temperature, voltage difference, timeout and recovery conditions of each candidate trajectory, eliminate the schemes with risks of overheating and overcooling, voltage difference exceeding the limit and charging timeout, and output the executable trajectory and balancing command. S4. Drive the power and equalization branch according to the target current and equalization command, sample and read back in real time, and adaptively switch between small-step fine-tuning and trying to cross the level according to the improvement rate and group diversity. After crossing the level, recalculate the subordinate parameters and location labels synchronously. If overcurrent, short circuit, overtemperature, timeout and voltage difference abnormality are encountered, immediately cut off and reduce the rate according to the graded strategy, and record the event for subsequent model fine-tuning and guardrail correction. S5. When the preset time is reached and the improvement is insufficient for multiple consecutive cycles, and the hibernation condition is triggered, the current cycle ends. The strategy that passes all hard boundary checks and has the best comprehensive score is selected as the final result. The equalization record, power and temperature, time curve, key event statistics and energy efficiency assessment are output simultaneously to provide a basis for subsequent cycle power-on initialization and remote maintenance. S1 specifically refers to: S1-1. During power-on and initial operation, acquire individual and total voltage, current, temperature and timing information according to the preset sampling cycle, complete range and zero drift correction, and perform noise reduction and outlier removal; combine historical samples and factory parameters to give initial estimates of SOC, SOH, equivalent internal resistance and key thermal parameters, generate an integrated dynamic model of electrical, thermal and aging, and mark the version number, timestamp and applicable temperature zone as the baseline for subsequent guardrail generation and trajectory planning; S1-2. During closed-loop execution, continuously receive readback data and event tags, including over-temperature, over-current, timeout, and equalization trigger. Perform small-step rolling correction and temperature-segmented correction on model parameters. Track the voltage and temperature differences between cells to assess consistency and degradation trends. Set isolation markers and weight reduction for channels with drift anomalies and sudden noise increases, and synchronously publish the updated state variables and boundary predictions to subsequent stages for use. S2 specifically refers to: S2-1. Without changing the hard thresholds and recovery conditions for overcharge, over-discharge, overcurrent, short circuit, and temperature specified in the datasheet, generate warning bands and hysteresis windows near the boundaries for key variables, and clearly define the allowable power range and recovery path; align the timing and alarm rules with this range to form an executable and implementable charge / discharge power-temperature joint constraint package, and output it for trajectory planning. S2-2. Combining ambient temperature, current State of Health (SOH), consistency level, and load scenario, the allowable range is adaptively shrunk and widened according to the scenario, and the start / stop thresholds of the balancing strategy and heat dissipation strategy are linked; a buffer zone is added for conditions near the boundary and the safety margin is recorded, outputting the final available power range and strategy trigger threshold, reducing the possibility of boundary back-and-forth triggering and false recovery, when the power change value Less than the lower critical value If this occurs, the scenario is defined as a potentially low safety margin scenario, and the voltage rebound amplitude is collected. The variance is calculated, if the voltage rebound amplitude variance Greater than the set value If so, it is determined to be in an overcharge protection state; S3 specifically refers to: S3-1. Within the allowable range, generate the target current trajectory using a rolling time window, constrain the current change rate, and link it with equalization as a co-variable to bring the voltage difference between cells closer to the allowable range. For different scenarios, including fast charging, low-temperature start-up and discharge, and long-term float charging, provide differentiated trajectory templates and initial values ​​to ensure improved energy utilization and temperature control efficiency without compromising safety. Locate the regions corresponding to high polarization and high internal resistance zones in the cell group state distribution map and electrochemical characteristic map, and increase the equalization trigger interval in these regions. The magnitude of the increase and and The ratios are positively correlated, specifically: ,in The default trigger interval is... This is the weighting coefficient for the voltage rebound amplitude and rate; S3-2. Verify the temperature, pressure difference, timing and recovery conditions of each candidate trajectory, and eliminate schemes with risks of overheating and overcooling, excessive pressure difference and charging timeout; automatically reduce the derating of schemes close to the boundary and provide backup trajectories, and finally output the executable main trajectory and equalization command, with execution priority and rollback conditions to ensure stable landing under field fluctuations and measurement disturbances. S4 specifically refers to: S4-1: Drive the power and equalization branch according to the target current and equalization command, compare the target and readback in real time and maintain within a small error range; when the improvement rate is good and far from the boundary, enter small-step refinement, reduce the disturbance amplitude to reduce costs and increase safety margin; when continuous improvement is insufficient and close to the upper edge, enter the attempt to cross the level, apply adsorption reward, amplify the disturbance amplitude and limit the exploration to a better direction, and immediately recalculate the subordinate parameters and location labels after the transition to maintain stability; S4-2. When overcurrent, short circuit, overtemperature, timeout and voltage difference abnormalities occur, quickly cut off and reduce the derating according to the graded strategy, and record the number of events, duration and recovery path; after recovery, make small-step corrections to the relevant thresholds and model parameters, and mark the key segments and equalization effects of this round of execution as inheritable segments, as the empirical basis for the next generation initial solution and subsequent batch initialization. S5 specifically refers to: S5-1. When the preset time and loop count limit are reached, and there are insufficient improvements within a certain number of consecutive time windows, and the condition for entering hibernation is met, the current optimization and execution loop is terminated. Before termination, a quick feasibility review and data integrity check are performed to ensure that the output results can be reproduced and audited. S5-2 outputs the final charge / discharge strategy, equalization execution record, temperature-time and voltage-time curves, key event statistics and energy efficiency assessment, and marks the applicable temperature range and health status range; at the same time, it writes to memory and remote logs, and provides initialization suggestions and parameter starting points for the next power-on, so as to be reused and maintained across batches and platforms.

[0017] This invention proposes a dynamic mathematical model-driven charge and discharge control system: under the premise of meeting the existing multi-level protection and temperature, equalization and timing requirements in the specification, online parameter identification and state estimation are introduced to construct an integrated dynamic model of the cell pack's electrical-thermal-aging, generate the optimal trajectory of charge and discharge power and equalization coordination according to the model prediction, and use multi-layer safety guardrails and double threshold hysteresis to suppress boundary jitter, and finally achieve closed-loop control that maximizes efficiency under safety and life constraints; Without altering the existing hard threshold system, prediction and soft guardrails are added to significantly reduce oscillations and false touches under critical conditions. The charging current and balancing current are dynamically allocated to reduce overcharging, overheating shutdowns, and deep over-discharge, thereby improving effective capacity utilization and cycle life. Threshold fine-tuning and trajectory optimization are driven by models and historical statistics, and are adaptive to cell consistency and aging status. The maintenance cost is low, and it is portable across platforms.

[0018] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another entity or operation, and do not necessarily imply any such actual relationship or order between these entities and operations. Furthermore, the terms "comprising," "including," and any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, and includes elements inherent to such a process, method, article, or apparatus.

[0019] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are 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 and 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 charging and discharging control system based on a dynamic mathematical model, characterized in that: The system includes a modeling module, a prediction and strategy module, and an execution and guardrail module. The modeling module is used to collect voltage, current, temperature, and time information, estimate the state of charge, health status, and thermo-electrical parameters online, and form a dynamic electrical, thermal, and aging model. The prediction and strategy module generates a rolling optimization trajectory for charging and discharging power and equalization coordination based on the model and safety guardrail, and outputs the target current, equalization command, and temperature management strategy. The execution and guardrail module drives the charging and discharging channels and equalization branches according to the target command, performs graded overcharge, over-discharge, overcurrent, short circuit, and temperature protection, and timeout alarms, records the status, and feeds back to the model.

2. The charging and discharging control system based on a dynamic mathematical model according to claim 1, characterized in that: The modeling module includes a parameter acquisition module, a state estimation module, and a model update module. The parameter acquisition module is connected to the cell assembly and temperature sensor via a signal to acquire information on individual cell and total voltage, current, temperature, and timing. The state estimation module is connected to the parameter acquisition module via a signal to estimate SOC, SOH, equivalent internal resistance, and thermal parameters. The model update module is connected to the state estimation module via a signal to perform rolling updates of model parameters based on historical records and alarm counts. The prediction and strategy module includes a guardrail generation module, a trajectory planning module, and a power allocation module. The guardrail generation module is connected to the model update module via a signal and is used to generate soft guardrails and hysteresis windows outside a fixed threshold. The trajectory planning module is connected to the guardrail generation module and the state estimation module via a signal and is used to plan the charging and discharging current and the equalization current in a rolling manner within constraints. The power allocation module is connected to the trajectory planning module via a control and is used to convert the trajectory into real-time instructions for charging, discharging, and equalization. The execution and guardrail module includes a power execution module, a protection control module, and an indication and memory module. The power execution module is electrically connected to the charging and discharging channel and the equalization branch, and is used to output according to instructions and sample and read back in real time. The protection control module is electrically connected to the measurement front end and the power switch, and is used to execute graded overcharge, over-discharge, overcurrent, short circuit and temperature protection and recovery strategies. The indication and memory module is connected to the protection control module and the model update module via signals, and is used for power indication, alarm display, memory of state before and after sleep mode and count of times.

3. A charging and discharging control method based on a dynamic mathematical model, characterized in that: The method includes the following steps: S1. During the startup phase, collect individual and total voltage, current, temperature and timing information, combine historical samples to obtain initial estimates of SOC, SOH and equivalent internal resistance and thermal parameters, generate an integrated dynamic model of electrical, thermal and aging. During operation, use readback data and event records, including overcurrent, overtemperature, timeout, and equalization triggers, to perform rolling correction of model parameters, output the current health status and prediction boundary as input for subsequent guardrail and trajectory planning. S2. In addition to the overcharge, over-discharge, overcurrent, short circuit and temperature hard thresholds specified in the specifications, an early warning zone and hysteresis window are automatically generated near the boundary to form the allowable power range that can be executed under the current operating conditions. The range is then shrunk and widened according to the scenario based on the ambient temperature, aging degree and differential pressure, and linked with the timing and recovery logic to reduce boundary back-and-forth triggering and false recovery. S3. Generate the target current trajectory for charging and discharging within the allowable range, and use balancing as a cooperative variable to make the voltage difference between cells tend to converge without affecting the main power path. Check the temperature, voltage difference, timeout and recovery conditions of each candidate trajectory, eliminate the schemes with risks of overheating and overcooling, voltage difference exceeding the limit and charging timeout, and output the executable trajectory and balancing command. S4. Drive the power and equalization branch according to the target current and equalization command, sample and read back in real time, and adaptively switch between small-step fine-tuning and trying to cross the level according to the improvement rate and group diversity. After crossing the level, recalculate the subordinate parameters and location labels synchronously. If overcurrent, short circuit, overtemperature, timeout and voltage difference abnormality are encountered, immediately cut off and reduce the rate according to the graded strategy, and record the event for subsequent model fine-tuning and guardrail correction. S5. When the preset time is reached and there are insufficient improvements in multiple consecutive cycles, and the hibernation condition is triggered, the current cycle ends. The strategy that passes all hard boundary checks and has the best comprehensive score is selected as the final result. The equalization record, power and temperature, time curve, key event statistics and energy efficiency assessment are output simultaneously to provide a basis for subsequent cycle power-on initialization and remote maintenance.

4. The charging and discharging control method based on a dynamic mathematical model according to claim 3, characterized in that: Specifically, S1 is: S1-1. During power-on and initial operation, acquire individual and total voltage, current, temperature and timing information according to the preset sampling cycle, complete range and zero drift correction, and perform noise reduction and outlier removal; combine historical samples and factory parameters to give initial estimates of SOC, SOH, equivalent internal resistance and key thermal parameters, generate an integrated dynamic model of electrical, thermal and aging, and mark the version number, timestamp and applicable temperature zone as the baseline for subsequent guardrail generation and trajectory planning; S1-2. During closed-loop execution, continuously receive readback data and event tags, including over-temperature, over-current, timeout, and equalization trigger, perform small-step rolling correction and temperature segment correction on model parameters, and track the voltage and temperature differences between cells to assess consistency and degradation trends. For channels with abnormal drift and sudden noise increases, isolation markers and weight reduction processing are set, and the updated state variables and boundary predictions are simultaneously published to subsequent processes.

5. The charging and discharging control method based on a dynamic mathematical model according to claim 4, characterized in that: Specifically, S2 is: S2-1. Without changing the hard thresholds and recovery conditions for overcharge, over-discharge, overcurrent, short circuit, and temperature specified in the datasheet, generate warning bands and hysteresis windows near the boundaries for key variables, and clearly define the allowable power range and recovery path; align the timing and alarm rules with this range to form an executable and implementable charge / discharge power-temperature joint constraint package, and output it for trajectory planning. S2-2. Combining ambient temperature, current State of Health (SOH), consistency level, and load scenario, the allowable range is adaptively shrunk and widened according to the scenario, and the start / stop thresholds of the balancing strategy and heat dissipation strategy are linked; a buffer zone is added for conditions near the boundary and the safety margin is recorded, outputting the final available power range and strategy trigger threshold, reducing the possibility of boundary back-and-forth triggering and false recovery, when the power change value Less than the lower critical value If this occurs, the scenario is defined as a potentially low safety margin scenario, and the voltage rebound amplitude is collected. The variance is calculated, if the voltage rebound amplitude variance Greater than the set value If so, it is determined to be in an overcharge protection state.

6. The charging and discharging control method based on a dynamic mathematical model according to claim 5, characterized in that: Specifically, S3 is: S3-1. Within the allowable range, generate the target current trajectory using a rolling time window, constrain the current change rate, and link it with equalization as a co-variable to bring the voltage difference between cells closer to the allowable range. For different scenarios, including fast charging, low-temperature start-up and discharge, and long-term float charging, provide differentiated trajectory templates and initial values ​​to ensure improved energy utilization and temperature control efficiency without compromising safety. Locate the regions corresponding to high polarization and high internal resistance zones in the cell group state distribution map and electrochemical characteristic map, and increase the equalization trigger interval in these regions. The magnitude of the increase and and The ratios are positively correlated, specifically: ,in The default trigger interval is... The weighting coefficients for voltage rebound amplitude and rate; S3-2. Verify the temperature, differential pressure, timing and recovery conditions of each candidate trajectory, and eliminate schemes with risks of overheating or overcooling, differential pressure exceeding limits and charging timeout; automatically reduce the derating of schemes close to the boundary and provide backup trajectories, and finally output the executable main trajectory and equalization command, with execution priority and rollback conditions to ensure stable landing under field fluctuations and measurement disturbances.

7. The charging and discharging control method based on a dynamic mathematical model according to claim 6, characterized in that: Specifically, S4 is: S4-1: Drive the power and equalization branch according to the target current and equalization command, compare the target and the readback in real time and maintain within a small error range; when the improvement rate is good and far from the boundary, enter small-step fine-tuning, reduce the disturbance amplitude to reduce costs and increase safety margin; When continuous improvement is insufficient and the target is close to the upper limit, an attempt is made to jump to a higher level. An adsorption reward is applied, the perturbation amplitude is amplified, and the exploration is limited to a better direction. At the same time, the subordinate parameters and location tags are recalculated immediately after the jump to maintain stability. S4-2. When overcurrent, short circuit, overtemperature, timeout and voltage difference abnormalities occur, quickly cut off and reduce the derating according to the graded strategy, and record the number of events, duration and recovery path; after recovery, make small-step corrections to the relevant thresholds and model parameters, and mark the key segments and equalization effects of this round of execution as inheritable segments, as the empirical basis for the next generation initial solution and subsequent batch initialization.

8. The charging and discharging control method based on a dynamic mathematical model according to claim 7, characterized in that: Specifically, S5 is: S5-1. When the preset time and loop count limit are reached, and there are insufficient improvements within a certain number of consecutive time windows, and the condition for entering hibernation is met, the current optimization and execution loop is terminated. Before termination, a quick feasibility review and data integrity check are performed to ensure that the output results can be reproduced and audited. S5-2 outputs the final charge / discharge strategy, equalization execution record, temperature-time and voltage-time curves, key event statistics and energy efficiency assessment, and marks the applicable temperature range and health status range; at the same time, it writes to memory and remote logs, and provides initialization suggestions and parameter starting points for the next power-on, so as to be reused and maintained across batches and platforms.

Images