A dynamic control method of a lithium battery module intelligent equalization management system

Abstract

The application relates to the technical field of battery management, and particularly discloses a dynamic control method of a lithium battery module intelligent equalization management system, which realizes real-time collection of load current and single-end voltage signals of a battery module; realizes real-time calculation of instantaneous internal resistance and polarization voltage parameters of each single battery; realizes real-time compensation of the end voltage to generate a purified open-circuit voltage signal; fuses a charge state reference value and a current integral result, and dynamically calculates a confidence index according to a current static condition and a voltage stability to realize self-adaptive weighting, and outputs a smooth charge state estimation value; combines the estimation value, real-time temperature and an inconsistency change trend to calculate an equalization urgency index to generate a differentiated equalization control instruction; establishes and controls an energy transfer path according to the instruction, dynamically adjusts the speed in the process, and terminates the operation when a time length or a state consistency condition is met; and the application improves the stability of state estimation, the accuracy, safety and efficiency of equalization control under dynamic working conditions.

Description

Technical Field

[0001] This invention relates to the field of battery management technology, and more specifically to a dynamic control method for an intelligent equalization management system for lithium battery modules. Background Technology

[0002] With the rapid development of electric vehicles and large-scale energy storage, lithium battery modules are widely used due to their high energy density and long cycle life. However, due to differences in manufacturing processes, operating environments, and cycle aging, the charge state of individual cells within a module inevitably becomes inconsistent, leading to a decrease in overall usable capacity, premature local overcharging / over-discharging, and even the risk of thermal runaway. Therefore, battery equalization management systems have become a key technology. Most existing equalization management methods are based on battery terminal voltage or simple ampere-hour integration for state estimation and decision-making. However, under actual dynamic operating conditions (such as low temperature and high-rate charge / discharge), the internal polarization effect of the battery is significant, and the terminal voltage signal will produce severe lag and distortion, causing jumps in the estimated charge state value. This, in turn, leads to misjudgment and oscillation of the equalization system, severely restricting the accuracy and reliability of equalization management and the overall performance and safety of the battery pack.

[0003] This invention aims to solve the key technical problem of severe distortion in charge state estimation caused by the time-varying characteristics of battery polarization voltage under extreme dynamic conditions such as low temperature and high-rate pulse, which leads to high-frequency oscillation of equalization control commands, ineffective energy transfer, and system instability. Specifically, traditional methods rely on terminal voltage or Kalman filtering of a fixed model for state estimation, which cannot isolate the dynamic polarization voltage component caused by drastic changes in load current in real time. This results in the estimation results fluctuating drastically with voltage lag. When the equalization system makes decisions based on this fluctuating signal, it leads to frequent switching of charge / discharge equalization modes for the same battery. This not only fails to achieve effective equalization but also accelerates the aging of power devices, increases unnecessary energy loss, and may induce safety accidents. Summary of the Invention

[0004] The purpose of this invention is to provide a dynamic control method for an intelligent equalization management system for lithium battery modules, so as to solve the problems mentioned above.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A dynamic control method for an intelligent balancing management system for lithium battery modules includes the following steps:

[0007] S1: Real-time acquisition of the operating status signals of each individual battery cell in the battery module, including load current signal and terminal voltage signal;

[0008] S2: Based on the load current signal and the terminal voltage signal, perform dynamic parameter online identification operation to calculate and obtain the instantaneous internal resistance parameter and instantaneous polarization voltage parameter of the single cell in real time;

[0009] S3: Using instantaneous internal resistance parameters and instantaneous polarization voltage parameters, the acquired terminal voltage signal is dynamically compensated to generate a purified open-circuit voltage signal corresponding to a single cell.

[0010] S4: The charge change obtained by integrating the purified open-circuit voltage signal with the load current signal is fused and calculated, and the charge state fusion calculation operation is performed to output the smoothed charge state estimate of the single cell.

[0011] S5: Based on the smoothed state of charge estimate, calculate the state of charge difference between individual cells in the battery module, and generate corresponding equalization control commands based on the result of the state of charge difference calculation.

[0012] S6: According to the equalization control command, control the equalization execution circuit to perform energy redistribution operation on the battery module to achieve intelligent equalization management.

[0013] As a further aspect of the present invention: S2 specifically includes:

[0014] S21, Construct a sliding time window with a preset duration, and dynamically buffer the acquired load current signal and terminal voltage signal within the sliding time window to obtain the current signal sequence and voltage signal sequence;

[0015] S22, based on the current signal sequence and voltage signal sequence, construct a set of nonlinear equations about the battery internal resistance parameter and polarization parameter;

[0016] S23 performs real-time solution of the nonlinear equation system and synchronously outputs the numerical solutions of the instantaneous internal resistance parameter and instantaneous polarization voltage parameter of the single cell at the current sampling time.

[0017] As a further aspect of the present invention: S3 specifically includes:

[0018] S31, based on the instantaneous internal resistance parameter and the real-time acquired load current signal, calculate the first real-time compensation component, and at the same time determine the second real-time compensation component based on the instantaneous polarization voltage parameter.

[0019] S32, dynamically calculate the dynamic weighting factor based on the rate of change of the load current signal, and use the dynamic weighting factor to weight and fuse the first real-time compensation component and the second real-time compensation component to generate a comprehensive compensation amount.

[0020] S33 subtracts the comprehensive compensation amount from the real-time acquired terminal voltage signal and outputs the compensation result as the purified open-circuit voltage signal.

[0021] As a further aspect of the present invention: S4 specifically includes:

[0022] S41, based on the purified open-circuit voltage signal, query the preset voltage-charge state correspondence table to obtain the first charge state reference value, and at the same time, obtain the second charge state reference value by performing compensated integral operation on the load current signal.

[0023] S42, based on the amplitude variation characteristics of the load current signal and the stability of the purification open-circuit voltage signal, dynamically calculate the confidence index of the first charge state reference value;

[0024] S43, construct an adaptive smoothing factor based on the confidence index, and use the smoothing factor to weight and fuse the first charge state reference value and the second charge state reference value to output a smoothed charge state estimate.

[0025] As a further aspect of the present invention: S42 specifically includes:

[0026] S421, calculate the absolute amplitude of the load current signal at the current time and within the previous preset time period. When the absolute amplitude of the number of consecutive hours exceeding the first threshold is lower than the static current threshold, determine that the load current signal is in a static state and generate the first determination factor; otherwise, generate the second determination factor.

[0027] S422, monitor the instantaneous fluctuation of the purification open-circuit voltage signal and count the cumulative duration of the instantaneous fluctuation being lower than the voltage stability threshold. When the cumulative duration exceeds the second threshold, determine that the purification open-circuit voltage signal is in a stable state and generate a third determination factor; otherwise, generate a fourth determination factor.

[0028] S423, based on the combination relationship between the first or second decision factor and the third or fourth decision factor, directly output the corresponding confidence index value through a pre-established confidence mapping table.

[0029] As a further aspect of the present invention: the construction of the adaptive smoothing factor based on the confidence index specifically includes:

[0030] The confidence index is input into a transformation function with a nonlinear mapping relationship. The transformation function maintains the minimum output value when the confidence index is below the first critical value, maintains the maximum output value when the confidence index is above the second critical value, and increases nonlinearly between the two critical values. The output value of the transformation function is used as the basic smoothing factor.

[0031] As a further aspect of the present invention: S5 specifically includes:

[0032] S51, calculate the average value of the smoothed state of charge estimates of all individual cells in the battery module, and calculate the deviation of the smoothed state of charge estimates of each individual cell from the average value.

[0033] S52 combines the deviation with the real-time temperature data of individual cells and calculates the balance urgency index of each individual cell through a pre-established balance urgency evaluation rule.

[0034] S53, based on the urgency index, query the equilibrium mode decision table to determine the corresponding equilibrium action instructions, including the equilibrium target object, equilibrium direction, and equilibrium intensity level.

[0035] As a further aspect of the present invention: the calculation process of the equilibrium urgency index is as follows:

[0036] The deviation was divided into three levels, and the real-time temperature data was divided into three activity levels. A basic mapping table of urgency containing nine assessment units was established.

[0037] Based on the level range of the deviation and the activity level range of the real-time temperature data, the corresponding assessment unit is determined, and the basic stress value is read from the stress basic mapping table.

[0038] Monitor the continuous trend of deviation. When the deviation changes continuously in the direction of increase for more than a preset period, add a trend enhancement factor to the basic urgency value to generate a balanced urgency index.

[0039] As a further aspect of the present invention: S6 specifically includes:

[0040] S61, based on the balance target object and balance direction in the balance control command, establish a directional energy transfer path from the source battery to the target battery, and determine the initial energy transfer rate based on the balance intensity level.

[0041] S62 monitors the source battery voltage change rate and target battery voltage change rate in real time during the energy transfer process. When either change rate exceeds the safety threshold, the energy transfer rate is dynamically reduced.

[0042] S63, when the energy transfer reaches the preset time or the difference between the smooth state of charge estimate of the target battery and the average value of the battery module enters the target range, the current energy redistribution operation is terminated.

[0043] The beneficial effects of this invention are:

[0044] (1) Traditional methods, under low temperature and high-rate pulse loads, cause severe fluctuations in terminal voltage due to battery polarization, leading to jumps in voltage-based SOC estimation results and causing malfunctions in the equalization system. This invention identifies the instantaneous ohmic internal resistance and polarization voltage parameters of the battery in real time online, and dynamically weights and fuses them based on the load current change rate to accurately compensate the terminal voltage signal dynamically, generating a "purified" open-circuit voltage signal. This signal effectively removes the dynamic voltage drop component caused by the operating conditions. Subsequently, combined with the current integration result, and based on the current resting condition and voltage stability, a confidence index is dynamically calculated for adaptive weighted fusion, ultimately outputting a smooth SOC estimate. This series of operations ensures that the SOC estimate is no longer affected by voltage transient lag, thus providing a reliable and stable state input for equalization decisions, eliminating the "equalization oscillation" phenomenon caused by signal jumps, and ensuring the reliability of system decisions under complex operating conditions.

[0045] (2) Based on accurate SOC estimation, the system calculates the differences between individual cells and further integrates real-time temperature data with the trend of inconsistency (monotonically increasing deviation). Through a preset urgency mapping table and trend enhancement factor, a comprehensive balancing urgency index is calculated. The balancing urgency index can distinguish between different demand levels such as "maintenance balancing" and "corrective balancing," thereby generating differentiated balancing instructions (including target, direction, and intensity). At the execution level, the system not only initializes the balancing path and rate according to the instructions, but also monitors the voltage change rate of the source battery and the target battery in real time during the energy transfer process. Once the safety threshold is exceeded, the rate is dynamically reduced, realizing closed-loop safety control of the process. The termination condition combines the preset duration and the consistency of the target battery state to avoid over-balancing or under-balancing. This intelligent management of the entire process from "multi-dimensional state assessment" to "adaptive decision-making" to "closed-loop safety execution" makes the balancing action more accurate, timely, and safe, effectively suppressing the expansion of battery inconsistency, optimizing energy utilization efficiency, and helping to extend the overall cycle life of the battery module. Attached Figure Description

[0046] The invention will now be further described with reference to the accompanying drawings.

[0047] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0048] 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.

[0049] Please see Figure 1 As shown, this invention provides a dynamic control method for an intelligent equalization management system for lithium battery modules, comprising the following steps:

[0050] S1: Real-time acquisition of the operating status signals of each individual battery cell in the battery module, including load current signal and terminal voltage signal;

[0051] S2: Based on the load current signal and the terminal voltage signal, perform dynamic parameter online identification operation to calculate and obtain the instantaneous internal resistance parameter and instantaneous polarization voltage parameter of the single cell in real time;

[0052] S3: Using instantaneous internal resistance parameters and instantaneous polarization voltage parameters, the acquired terminal voltage signal is dynamically compensated to generate a purified open-circuit voltage signal corresponding to a single cell.

[0053] S4: The charge change obtained by integrating the purified open-circuit voltage signal with the load current signal is fused and calculated, and the charge state fusion calculation operation is performed to output the smoothed charge state estimate of the single cell.

[0054] S5: Based on the smoothed state of charge estimate, calculate the state of charge difference between individual cells in the battery module, and generate corresponding equalization control commands based on the result of the state of charge difference calculation.

[0055] S6: According to the equalization control command, control the equalization execution circuit to perform energy redistribution operation on the battery module to achieve intelligent equalization management.

[0056] In S1, the load current signal is acquired through a Hall current sensor deployed in the main circuit of the battery module. This sensor converts the current flowing through the entire module into a proportional analog voltage signal, which is then sent to the analog-to-digital converter of the battery management system to be converted into a digital load current signal.

[0057] The acquisition of the terminal voltage signal is performed individually for each single cell. For each single cell, its positive and negative terminals are connected to a high-precision voltage detection circuit via wires. This circuit typically includes a precision voltage divider network and an isolated operational amplifier, which can linearly and safely convert the battery's terminal voltage to the measurable range of the analog-to-digital converter, thereby obtaining the real-time terminal voltage signal of each single cell.

[0058] In step S21 of S2, the construction of the sliding time window is specifically implemented as follows: a fixed-length data buffer is allocated for each individual battery cell in the memory of the battery management system. This buffer continuously stores historical signal data for a preset duration preceding the previous sampling time. The preset duration is determined based on the battery polarization dynamic response time and is set to 2 to 5 seconds, for example, 3 seconds. At a sampling frequency of 100 Hz, this buffer stores approximately 300 consecutive historical sampling points. At each new sampling time, the oldest historical data point is removed from the buffer, and the new load current signal and the corresponding terminal voltage signal of the individual battery cell are stored, thereby forming and continuously updating a current signal sequence and voltage signal sequence of length 300 points. This process achieves the preservation of recent dynamic data.

[0059] In step S22, the nonlinear equation set is constructed based on a first-order equivalent circuit model. This model describes the internal state of a single cell at a certain moment as follows: the terminal voltage at that moment is equal to the sum of its open-circuit voltage, the ohmic internal resistance voltage drop, and the polarization voltage. The ohmic internal resistance voltage drop is obtained by multiplying the load current at that moment by an instantaneous ohmic internal resistance parameter; the dynamic change of the polarization voltage is described by a parallel resistor-capacitor branch, and its rate of change is related to the load current and the polarization voltage itself, involving two parameters: instantaneous polarization resistance and instantaneous polarization capacitance. To utilize discrete current and voltage signal sequences, the above continuous relationship is transformed into discrete difference equations. Specifically, for the two sets of sequences buffered within the sliding time window, starting from the first data point, an equation can be written between every two adjacent sampling points based on this difference relationship. For a sequence of length 300, a system of equations containing nearly 300 equations can be constructed, where the unknowns are the instantaneous open-circuit voltage, instantaneous ohmic internal resistance parameter, instantaneous polarization resistance parameter, and instantaneous polarization capacitance parameter at each sampling time. Since the open-circuit voltage changes slowly, it can be approximated as a constant within a 3-second window, thus simplifying the system of equations into a form where the constant open-circuit voltage and the aforementioned three dynamic parameters are the main unknowns.

[0060] In step S23, the real-time solution is implemented using a recursive solution method with a forgetting factor. At the initial moment, a set of reasonable preset initial values ​​are assigned to the unknown parameter vector to be solved, and a covariance matrix is ​​initialized. When a new sampling moment arrives and the latest current and voltage data are obtained, the solution process proceeds as follows: First, based on the parameter estimates from the previous moment, the current terminal voltage value is predicted using discrete difference equations. Next, the difference between the predicted value and the actually acquired terminal voltage signal is calculated. Subsequently, based on this difference, the current load current signal, and a gain vector updated over time, the parameter estimate vector from the previous moment is corrected to obtain the parameter estimate value for the current moment. The calculation of the gain vector depends on the covariance matrix, the load current signal, and a forgetting factor between 0.95 and 0.995. This forgetting factor helps the algorithm focus more on recent data, thereby tracking the time-varying characteristics of the parameters. Finally, the covariance matrix is ​​updated to prepare for the calculation at the next moment. Through the above recursive steps, a set of corresponding instantaneous ohmic internal resistance parameter numerical solutions can be output at each sampling time, as well as instantaneous polarization voltage parameter numerical solutions calculated based on the current polarization parameter estimate and the load current history.

[0061] In step S3, specifically step S31, the determination of the first and second real-time compensation components is implemented as follows: The first real-time compensation component is calculated by multiplying the value of the instantaneous internal resistance parameter (specifically, the instantaneous ohmic internal resistance parameter) obtained in step S23 at the current sampling time with the value of the load current signal acquired in real-time at the same sampling time in step S1. This product physically represents the real-time voltage drop caused by the battery's ohmic internal resistance. The second real-time compensation component is determined by directly using the numerical solution of the instantaneous polarization voltage parameter calculated and output in step S23. This numerical solution physically characterizes the real-time voltage component caused by the electrochemical polarization process inside the battery. Through the above calculation and direct application, two independent compensation quantities corresponding to the ohmic effect and polarization effect are obtained simultaneously.

[0062] In step S32, the calculation of the dynamic weighting factor and the generation of the comprehensive compensation amount are specifically implemented as follows: First, the rate of change of the load current signal is calculated by taking the load current value at the current sampling moment and the load current value at the previous sampling moment, calculating the difference between the two, and then dividing the difference by the time interval between the two sampling points (i.e., the reciprocal of the sampling period). Next, the absolute value of the rate of change is judged to determine the dynamic weighting factor: a current change rate threshold is preset, for example, 10 times the rated capacity current per second. When the absolute value of the rate of change is lower than this threshold, the current change is determined to be gradual, and the dynamic weighting factor is set to a first fixed value, for example, 0.1; when the absolute value of the rate of change is higher than or equal to this threshold, the current change is determined to be drastic, and the dynamic weighting factor is set to a second fixed value, for example, 0.9. Then, a weighted fusion is performed to generate a comprehensive compensation amount: the second real-time compensation component (polarization voltage component) is weighted using the calculated dynamic weighting factor, and the first real-time compensation component (ohmic voltage drop component) is weighted using the value obtained by subtracting the dynamic weighting factor from the value "1". Finally, the two weighted results are added together. The calculation logic described in words is: the comprehensive compensation amount equals (dynamic weighting factor) multiplied by (second real-time compensation component), plus (one minus the dynamic weighting factor) multiplied by (first real-time compensation component).

[0063] In step S33, the generation of the purified open-circuit voltage signal is specifically implemented as follows: The value of the real-time terminal voltage signal of the individual battery corresponding to the current calculation time, acquired in step S1, is subtracted from the value of the comprehensive compensation amount calculated in step S32. The result of this subtraction is the voltage value after dynamic compensation, which is defined and output as the purified open-circuit voltage signal. Physically, this operation removes the voltage drop caused by the instantaneous load current through the internal resistance and the main dynamic polarization voltage components from the measured total terminal voltage, thereby obtaining an estimated signal that is closer to the true open-circuit voltage under battery equilibrium conditions, providing a more stable voltage reference for subsequent steps.

[0064] In step S41 of S4, the specific implementation of obtaining the first charge state reference value and the second charge state reference value is as follows: The first charge state reference value is obtained by querying a pre-stored voltage-charge state correspondence table. This table is obtained by experimentally determining the mapping relationship between the open-circuit voltage of a single cell at standard temperature and after sufficient rest and the corresponding charge state percentage, and then discretizing and storing it. Using the value of the purified open-circuit voltage signal output from step S3 as the query input, the corresponding charge state percentage is obtained through table lookup and linear interpolation. This value is the first charge state reference value, denoted as... The second charge state reference value is obtained by performing compensated integration on the load current signal acquired in step S1. The integration operation starts with a known initial charge state value. Within each sampling period, the load current value is multiplied by the sampling time interval and then divided by the nominal capacity of the battery to obtain the charge state change within that period, which is then accumulated. The compensation is reflected in the correction of the coulombic efficiency and the compensation for zero drift of the current sensor. Specifically, let the second charge state reference value be... The load current at each sampling time is (Discharge is positive), sampling interval is The nominal capacity is The Coulomb efficiency correction factor is (Taking values ​​from 0.98 to 0.998), the zero drift compensation value is... Then from the initial time to the th The second charge state reference value at each moment The textual description of the calculation process is as follows: initial value Given; for From 1 to ,calculate: ;

[0065] In step S42, the dynamic calculation of the confidence index of the first charge state reference value is implemented according to steps S421 to S423. In step S421, the determination of the quiescent state is implemented as follows: The quiescent current threshold is set to 0.05 times the nominal battery capacity, i.e., a current value of 0.05C; the first threshold duration is set to 5 seconds. The absolute amplitude of the load current signal corresponding to each sampling point within the current moment and the previous 4 seconds (assuming a sampling frequency of 1 Hz, then a total of 5 sampling points) is calculated. It is checked whether these 5 consecutive absolute amplitudes are all lower than the quiescent current threshold. If all are lower, the load current signal is determined to be in a quiescent state, and a first determination factor with a value of 1 is generated. Otherwise, it is determined to be in a non-static state, and a second determination factor with a value of 0 is generated. In step S422, the determination of the stable state is implemented as follows: the voltage stability threshold is set to 1 millivolt; the second threshold duration is set to 10 seconds. The purified open-circuit voltage signal is monitored, and the absolute value of the difference between two adjacent sampling points is calculated, defined as the instantaneous fluctuation. Statistical instantaneous fluctuations The cumulative duration of the voltage remaining below 1 millivolt. If this cumulative duration reaches or exceeds 10 seconds, the purified open-circuit voltage signal is determined to be in a stable state, and a third determination factor with a value of 1 is generated. Otherwise, generate the fourth determination factor with a value of 0. In step S423, the output of the confidence index value is achieved by querying a pre-established confidence mapping table. This mapping table is based on... and The confidence index is directly given by the combination relationship. Its mapping rule is defined as: when and , ;when and , ;when and , ;when and , The confidence index The higher the value, the more reliable the first charge state reference value obtained based on the clean open-circuit voltage signal is considered.

[0066] In step S43, the construction of the adaptive smoothing factor and the output of the smoothed charge state estimate are specifically implemented as follows. First, based on the confidence index... The adaptive smoothing factor is constructed. The transformation function is implemented using a three-segment nonlinear function, with the first critical value set to 0.4 and the second critical value set to 0.8. The output value of the transformation function is used as the basic smoothing factor. The calculation rule is: if If less than 0.4, then =0.1; if If it is greater than 0.8, then =0.9; if Between 0.4 and 0.8 ;

[0067] Subsequently, the base smoothing factor is fine-tuned to obtain the final adaptive smoothing factor. The fine-tuning process takes into account the instantaneous rate of change of the load current signal. The calculation method is to divide the absolute value of the difference between the current and the current at the previous sampling point by the sampling time interval. A rate of change threshold is set. It is twice the nominal current per second (i.e., 2C / second). If The calculation method is to divide the absolute value of the difference between the current and the current at the previous sampling point by the sampling time greater than 10 ... Then Set as Multiply by a decay factor less than 1 (e.g., 0.7); otherwise, Finally, an adaptive smoothing factor is used. Reference value for the first charge state Second charge state reference value Perform weighted fusion to output smoothed charge state estimates. The calculation formula is: ;

[0068] When the voltage observation signal is stable and the current is at rest, the smoothing factor Approaching 0.9, the fusion result mainly depends on the first charge state reference value. Under dynamic operating conditions and when the voltage is unstable, Approaching 0.1 or lower, the fusion result depends primarily on the second charge state reference value. The integral trajectory is obtained, thereby achieving adaptive smooth estimation and effectively suppressing the jump in charge state estimation caused by instantaneous signal fluctuations.

[0069] In S5, step S51, the calculation of the average value of the smoothed state of charge estimate and the corresponding deviation of each individual cell is specifically implemented as follows. First, assume that the battery module contains a total of Each individual cell, in the first... The smoothed state of charge estimate for each sampling time is output by step S4. The average state of charge of the battery module at that time is calculated.

[0070] In step S51, the calculation of the average value and deviation of the smoothed charge state estimate is specifically implemented as follows. First, assume the battery module consists of... The system consists of several individual cells connected in series. The smoothed state-of-charge estimate of each individual cell at the current sampling time is denoted as... subscript Representing the Individual battery cells, The value range is 1 to Calculate the arithmetic mean of the smoothed state-of-charge estimates for all individual cells, denoted as . The calculation process, as described in the text, is as follows: The values ​​of the first to mth individual cells... All values ​​are summed, and then the sum is divided by the total number of individual cells, m. This average value, SOCest > SOCest, characterizes the overall state of charge level of the battery module at the current moment. Subsequently, for each individual cell, its smoothed state of charge estimate is calculated. Compared with this average value The absolute value of the difference is the deviation of the individual cell, denoted as . That is, for the first... Individual cells: This deviation quantifies the degree of difference between the average state of charge of the individual cell and the module.

[0071] In step S52, the calculation of the equilibrium urgency index is specifically implemented as follows. First, the deviation degree... It is divided into three level ranges. The first level range is defined as follows: Less than 0.5%, the second level range is The third level range is equal to or less than 0.5% and less than 2%. Greater than or equal to 2%. Simultaneously, the real-time temperature data of individual battery cells will be... Divided into three activity level ranges: the low-temperature inhibition region is defined as... The normal operating range below 0 degrees Celsius is defined as follows: The high-temperature accelerated aging zone is defined as being greater than or equal to 0 degrees Celsius and less than or equal to 45 degrees Celsius. Above 45 degrees Celsius. Based on these two three-level divisions, a basic urgency mapping table is established, containing nine assessment units in three rows and three columns. The content of this mapping table is pre-defined; for example, when the deviation belongs to the first interval and the temperature belongs to the low-temperature inhibition zone, the basic urgency value is... Set to 0.1; when the deviation is in the first range and the temperature is in the normal temperature range, Set to 0.2; when the deviation is in the first range and the temperature is in the high-temperature range, Set to 0.3; when the deviation is in the second range and the temperature is in the low-temperature range, Set to 0.4; when the deviation is in the second range and the temperature is in the normal temperature range, Set to 0.6; when the deviation is in the second range and the temperature is in the high-temperature range, Set to 0.8; when the deviation is in the third interval and the temperature is in the low-temperature zone, Set to 0.7; when the deviation is in the third range and the temperature is in the normal temperature range, Set to 0.9; when the deviation is in the third interval and the temperature is in the high-temperature zone, Set to 1.0. During runtime, the deviation is calculated in real time based on the j-th individual cell. The corresponding grade range and its temperature The corresponding assessment unit is determined based on the activity level range, and the baseline urgency value is directly read from the mapping table. Next, monitor the deviation. The continuous changing trend. The preset period is set to 5 consecutive sampling periods. Calculate the current time... Compared with the previous 4 sampling times Historical values: If five consecutive values ​​show a monotonically increasing trend (i.e., each subsequent value is strictly greater than the previous value), then the deviation is determined to have increased continuously beyond a preset period. In this case, a trend reinforcement factor is generated. Its value is set to 0.2; otherwise, Set to 0. Ultimately, the equilibrium urgency index of this single cell is... It is obtained by adding the base urgency value to the trend reinforcement factor, i.e.: The aforementioned equilibrium urgency index The value ranges from 0.1 to 1.2, and the higher the value, the more urgent the need for equalization intervention on the individual cell.

[0072] In step S53, the determination of the balancing action command is specifically implemented as follows: A balancing mode decision table is pre-established, which uses a balancing urgency index... To query the index, three types of balancing action instructions were defined. Specifically, two thresholds were set: a low urgency threshold and a low urgency threshold. The threshold for high urgency is 0.3. It is 0.7. When a certain single cell... Below At that time, the determined balancing action is "do not perform balancing," meaning that the battery is not set as the balancing target object. Greater than or equal to and less than At this time, the determined balancing action is "perform maintenance balancing". In this case, the individual battery cell is set as the balancing target, and the balancing direction is determined according to its... Above or below the average The system determines the balancing direction: if the current is above average, the balancing direction is set to discharge, transferring energy out; if the current is below average, the balancing direction is set to charge, replenishing energy to the battery from external sources. The balancing intensity level is set to "low intensity" in this mode, corresponding to a lower balancing current, such as 0.05% of the nominal capacity (0.05C). Greater than or equal to At this time, the determined balancing action is "perform corrective balancing". In this case, the individual cell is also set as the balancing target. The balancing direction is determined by the same rules as for maintenance balancing, but the balancing intensity level is set to "high intensity", corresponding to a higher balancing current, such as 0.2% of the nominal capacity (0.2C). This is calculated for all individual cells in the battery module. By sequentially querying this decision table, a set of equalization control instructions can be generated for each individual cell requiring equalization, including the cell itself as the target object, the specific equalization direction, and the corresponding intensity level. This set of instructions will then be passed to subsequent execution steps.

[0073] In S6, step S61, the establishment of the directional energy transfer path and the determination of the initial energy transfer rate are specifically implemented as follows. First, the balancing control command from step S53 is parsed. The command contains a clear balancing target, namely the specific cell number that needs to receive energy replenishment or release energy. The source and target of energy transfer are determined according to the balancing direction: if the balancing direction is "charging", energy flows from the external common energy storage unit or the designated high-capacity battery (source battery) to the target battery; if the direction is "discharging", energy flows from the target battery (which is now the source battery) to the external common energy storage unit or the designated low-capacity battery (target battery). The process of establishing the path is manifested by closing the corresponding power switch combination, physically forming a conductive loop that allows current to flow from the positive terminal of the source battery, through the balancing inductor or transformer, and finally into the positive terminal of the target battery. Simultaneously, the initial energy transfer rate is determined based on the equalization intensity level in the instruction: if the intensity level is "low intensity," the duty cycle of the pulse width modulation signal in the equalization circuit is set to a first preset value, for example, a duty cycle corresponding to an equalization current of 0.05 times the battery's nominal capacity; if the intensity level is "high intensity," the duty cycle is set to a second preset value, for example, a duty cycle corresponding to an equalization current of 0.2 times the battery's nominal capacity. This duty cycle directly determines the initial average current of energy transfer, i.e., the initial energy transfer rate.

[0074] In step S62, the dynamic reduction of the energy transfer rate is implemented as follows: After the equalization circuit is activated and energy transfer begins, the terminal voltages of the source battery and the target battery are synchronously collected at a fixed monitoring cycle (e.g., 10 times per second). The instantaneous rate of change of the source battery terminal voltage and the instantaneous rate of change of the target battery terminal voltage are calculated within each monitoring cycle. The rate of change is calculated by taking the voltage sample value of the current cycle and the voltage sample value of the previous cycle, calculating the difference between the two, and then dividing the difference by the time interval between the two sampling cycles. A uniform safety threshold is set, for example, a voltage change of 0.1 volts per second. The source battery voltage change rate is compared with this safety threshold in real time, and the target battery voltage change rate is also compared with this safety threshold. If the value of either rate of change exceeds 0.1 volts per second, a dynamic reduction operation is triggered. The reduction is implemented by multiplying the duty cycle value of the pulse width modulation signal currently controlling the equalization circuit by an attenuation coefficient between 0 and 1 (e.g., 0.5), thereby immediately reducing the equalization current, i.e., reducing the energy transfer rate. After the voltage is reduced, continue to monitor the rate of change of voltage. If it still exceeds the safety threshold, repeat this attenuation operation until the rate of change is lower than the safety threshold or a minimum allowable duty cycle is reached.

[0075] In step S63, the termination of the energy redistribution operation is specifically implemented as follows: Two conditions are monitored in parallel during the equalization process. The first condition is the duration of energy transfer: the equalization time accumulated from the establishment of the current equalization path and the start of energy transfer. A preset maximum operation duration is set for this equalization, for example, 600 seconds for "maintenance equalization" (low intensity) and 150 seconds for "corrective equalization" (high intensity). When the accumulated equalization time reaches this preset duration, the operation is terminated immediately. The second condition is the state consistency of the target battery: the absolute value of the difference between the current smoothed state of charge estimate of the target battery (continuously updated by step S4) and the current average value of the battery module (continuously updated by step S51) is calculated in real time. A target range is set, for example, the absolute value of the difference is less than 0.1%. The difference is compared with 0.1% in real time, and the operation is terminated immediately once the difference is detected to enter the target range (i.e., less than 0.1%). These two conditions form an "OR" logical relationship; that is, if either condition is met, the transmission of pulse width modulation signals to the equalization circuit immediately stops, the corresponding power switch is disconnected, and the current directional energy redistribution ends. After the operation terminates, the equalization resources occupied by that path are released, and the system can respond to new equalization commands.

[0076] The working principle of this invention is as follows: First, the total load current signal of the battery module and the terminal voltage signal of each individual battery cell are acquired in real time. Second, based on the acquired signals, a historical data sequence within a sliding time window is constructed, and the instantaneous ohmic internal resistance and polarization voltage parameters of each individual battery cell are calculated in real time using a recursive algorithm with a forgetting factor. Next, the terminal voltage is compensated in real time by dynamically weighting and fusing the above parameters according to the load current change rate, generating a purified open-circuit voltage signal stripped of dynamic polarization effects. Then, the charge state reference value obtained from the table lookup of this purified voltage signal is fused with the reference value obtained from current integration, and the voltage is dynamically calculated based on the current resting state and voltage stability. The confidence index is derived from adaptive weighting, outputting a smoothed state of charge estimate. Subsequently, the deviation of each individual cell estimate from the module average is calculated. Combined with real-time temperature data and a preset mapping table and trend analysis, the urgency index for equalization is determined. Based on this, equalization control commands containing the target object, direction, and intensity level are generated. Finally, a directional energy transfer path is established according to the commands and executed at an initial rate. During the transfer process, the voltage change rate is monitored in real time to dynamically adjust the rate. The operation terminates when a preset time is reached or when the difference between the target cell state and the module average enters the target range, thereby achieving efficient, stable, and adaptive intelligent equalization management of the battery module.

[0077] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A dynamic control method of a lithium battery module intelligent equalization management system, characterized in that, The method comprises the following steps: S1: collecting the running state signals of each single battery in the battery module in real time, including the load current signal and the terminal voltage signal; S2: based on the load current signal and the terminal voltage signal, performing dynamic parameter online identification operation to solve and obtain the instantaneous internal resistance parameter and the instantaneous polarization voltage parameter of the single battery in real time; S3: using the instantaneous internal resistance parameter and the instantaneous polarization voltage parameter to perform dynamic compensation on the collected terminal voltage signal to generate a purified open circuit voltage signal corresponding to the single battery, specifically including: S31, based on the instantaneous internal resistance parameter and the real-time collected load current signal, a first real-time compensation component is calculated, and a second real-time compensation component is determined based on the instantaneous polarization voltage parameter; S32, a dynamic weight factor is dynamically calculated according to the change rate of the load current signal, and the first real-time compensation component and the second real-time compensation component are weighted and fused by using the dynamic weight factor to generate a comprehensive compensation; S33, the comprehensive compensation is subtracted from the real-time collected terminal voltage signal, and the compensation result is output as the purified open circuit voltage signal; S4: fusing and solving the purified open circuit voltage signal and the charge change amount obtained by integrating the load current signal to perform charge state fusion solving operation, and outputting the smooth charge state estimation value of the single battery, specifically including: S41, querying the preset voltage-charge state corresponding relationship table based on the purified open circuit voltage signal to obtain a first charge state reference value, and simultaneously obtaining a second charge state reference value by performing compensated integration operation on the load current signal; S42, dynamically calculating the confidence index of the first charge state reference value according to the amplitude change characteristics of the load current signal and the stability degree of the purified open circuit voltage signal; S43, constructing an adaptive smoothing factor based on the confidence index, and weighting and fusing the first charge state reference value and the second charge state reference value by using the smoothing factor to output the smooth charge state estimation value; S5: based on the smooth charge state estimation value, performing charge state difference calculation between the single batteries in the battery module, and generating corresponding equalization control instructions according to the charge state difference calculation result; S6: controlling the equalization execution circuit to perform energy redistribution operation on the battery module according to the equalization control instructions to realize intelligent equalization management.

2. The dynamic control method of the intelligent equalization management system of the lithium battery module according to claim 1, characterized in that, The S2 specifically includes: S21, constructing a sliding time window with a preset time length as a period, and dynamically caching the collected load current signal and terminal voltage signal in the sliding time window to obtain a current signal sequence and a voltage signal sequence; S22, based on the current signal sequence and the voltage signal sequence, constructing a nonlinear equation set about the battery internal resistance parameter and the polarization parameter; S23, real-time solving the nonlinear equation set to synchronously output the numerical solution of the instantaneous internal resistance parameter and the instantaneous polarization voltage parameter of the single battery at the current sampling time.

3. The dynamic control method of the intelligent equalization management system of the lithium battery module according to claim 1, characterized in that, The S4 specifically includes: S41, querying the preset voltage-charge state corresponding relationship table based on the purified open circuit voltage signal to obtain a first charge state reference value, and simultaneously obtaining a second charge state reference value by performing compensated integration operation on the load current signal; S42, dynamically calculate a confidence index of the first state of charge reference value according to a variation characteristic of the amplitude of the load current signal and a stability degree of the purified open circuit voltage signal; S43, construct an adaptive smoothing factor based on the confidence index, and perform weighted fusion on the first state of charge reference value and the second state of charge reference value by using the smoothing factor, and output a smoothed state of charge estimation value.

4. The dynamic control method of the intelligent equalization management system of the lithium battery module according to claim 1, characterized in that, The S42 specifically comprises: S421, calculate the absolute amplitude of the load current signal at the current moment and within a previous preset time period, when the absolute amplitudes continuously exceeding a first threshold value for a first threshold number of times are all lower than a static current threshold value, determine that the load current signal is in a static state and generate a first determination factor, otherwise generate a second determination factor; S422, monitor the instantaneous fluctuation amount of the purified open circuit voltage signal, and count the cumulative duration of the instantaneous fluctuation amount continuously being lower than a voltage stability threshold value, when the cumulative duration exceeds a second threshold number of times, determine that the purified open circuit voltage signal is in a stable state and generate a third determination factor, otherwise generate a fourth determination factor; S423, according to the combination relationship of the first determination factor or the second determination factor and the third determination factor or the fourth determination factor, directly output the corresponding confidence index value through a pre-established confidence mapping table.

5. The dynamic control method of the intelligent equalization management system of the lithium battery module according to claim 1, characterized in that, The adaptive smoothing factor is constructed based on the confidence index, and specifically comprises: input the confidence index into a conversion function with a nonlinear mapping relationship, wherein the conversion function maintains a minimum output when the confidence index is lower than a first critical value, maintains a maximum output when the confidence index is higher than a second critical value, and increases according to a nonlinear characteristic between the two critical values, and the output value of the conversion function is taken as a basic smoothing factor.

6. The dynamic control method of the intelligent equalization management system of the lithium battery module according to claim 1, characterized in that, The S5 specifically comprises: S51, calculate the average value of the smoothed state of charge estimation values of all single batteries in the battery module, and calculate the deviation degree of each single battery from the average value; S52, combine the deviation degree with the real-time temperature data of the single battery, and calculate the balancing urgency index of each single battery through a pre-established balancing urgency evaluation rule; S53, according to the balancing urgency index, query a balancing mode decision table to determine the corresponding balancing action instruction, including the balancing target object, the balancing direction and the balancing intensity level.

7. The dynamic control method of the intelligent equalization management system of the lithium battery module according to claim 6, characterized in that, The calculation process of the balancing urgency index is as follows: divide the deviation degree into three level intervals, and divide the real-time temperature data into three activity level intervals, and establish a urgency basic mapping table containing nine evaluation units; determine the corresponding evaluation unit according to the level interval to which the deviation degree belongs and the activity level interval to which the real-time temperature data belongs, and read the basic urgency value from the urgency basic mapping table; monitor the continuous change trend of the deviation degree, and when the deviation degree continuously changes in the increasing direction for more than a preset period, add a trend strengthening factor to the basic urgency value to generate the balancing urgency index.

8. The dynamic control method of the intelligent equalization management system of the lithium battery module according to claim 1, characterized in that, The S6 specifically comprises: S61, according to the balancing target object and the balancing direction in the balancing control instruction, establish a directional energy transfer path from the source battery to the target battery, and determine an initial energy transfer rate based on the balancing intensity level; S62, real-time monitoring of the source battery voltage rate of change and the target battery voltage rate of change in the energy transfer process, when any rate of change exceeds the safety threshold, dynamically reducing the energy transfer rate; S63, when the energy transfer reaches the preset time length or the difference between the smooth charge state estimation value of the target battery and the average value of the battery module enters the target interval, terminating the energy redistribution operation.

Images