Battery pack equalization control method and system based on multi-dimensional state mapping

Abstract

The invention discloses a battery pack equalization control method and system based on multidimensional state mapping, and the method comprises the steps: obtaining terminal voltage data, working current data and surface temperature data of each single battery in a target battery pack; calculating an ohm internal resistance value and a polarization impedance value of each single battery; calculating an impedance growth rate; according to the impedance growth rate and the surface temperature data, a health state value is determined, and an aging weighting coefficient is determined; determining a polarization safety coefficient of each single battery based on the polarization impedance value; calculating a state-of-charge value of each single battery in the target battery pack, and determining a balance demand deviation degree; determining a basic equalization duty ratio of each single battery according to the equalization demand deviation degree; correcting the basic balance duty ratio to generate a target pulse width modulation duty ratio; and according to the target pulse width modulation duty ratio, controlling equalization circuits connected to the two ends of each single battery to be switched on. The battery pack equalization control efficiency is improved.

Description

Technical Field

[0001] This application relates to the field of battery pack equalization control technology, specifically to a battery pack equalization control method and system based on multidimensional state mapping. Background Technology

[0002] With the rapid development of new energy vehicles and energy storage systems, lithium-ion batteries have become the mainstream energy storage and supply unit due to their advantages in energy density, cycle life, and safety. To meet the system's requirements for high voltage and large capacity, multiple individual cells are usually connected in series or parallel to form a battery pack. However, due to differences in manufacturing processes, inconsistent operating environments, and different aging rates, the individual cells in the battery pack will exhibit differences in performance parameters such as voltage, current, and internal resistance during cycle use. This can lead to a decline in the overall performance of the battery pack, a shortened lifespan, and even safety risks.

[0003] In existing technologies, active balancing redistributes energy among different individual cells through charge transfer. However, battery performance degradation during long-term operation is inconsistent, such as increased internal resistance or enhanced polarization. If the distribution of balancing current is determined solely by the voltage difference or state of charge difference between individual cells, applying a balancing current of conventional strength to an aging cell with high polarization resistance will generate severe polarization overpotential (false voltage high or low) across its terminals. This can lead to the Battery Management System (BMS) misjudging the battery voltage as reaching its limit and frequently triggering protection, forcing the interruption of the balancing process. Furthermore, the additional polarization heat generation may exacerbate internal side reactions within the battery, causing the balancing strategy to fail and reducing the efficiency of battery pack balancing control. Summary of the Invention

[0004] This application provides a battery pack equalization control method and system based on multidimensional state mapping, which solves the technical problem that relying solely on the voltage difference or state of charge difference of individual cells to determine the distribution of equalization current leads to forced interruption of the equalization process or failure of the equalization strategy, thereby improving the efficiency of battery pack equalization control.

[0005] The first aspect of this application provides a battery pack equalization control method based on multidimensional state mapping, the method comprising: Acquire the terminal voltage data, operating current data, and surface temperature data of each individual cell in the target battery pack; Construct an equivalent circuit model for each of the individual cells, and calculate the ohmic internal resistance and polarization impedance of each individual cell based on the terminal voltage data, the operating current data, and the equivalent circuit model. Calculate the impedance growth rate of the ohmic internal resistance value relative to the preset nominal internal resistance value; Based on the impedance growth rate and the surface temperature data, the health status value of each individual cell is determined, and an aging weighting coefficient is determined based on the health status value of each individual cell. The polarization safety factor of each individual cell is determined based on the polarization impedance value; The state of charge (SOC) value of each individual cell in the target battery pack is determined based on a preset charge calculation model, and the balance demand deviation is determined based on the SOC value. The basic balance duty cycle of each individual cell is determined based on the balance demand deviation. The basic equalization duty cycle is corrected based on the polarization safety factor and the aging weighting factor to generate the target pulse width modulation duty cycle; The equalization circuit connected to both ends of each individual cell is turned on according to the target pulse width modulation duty cycle.

[0006] Optionally, based on the terminal voltage data, the operating current data, and the equivalent circuit model, the ohmic internal resistance and polarization impedance of each individual cell are calculated, specifically including: Monitor the time change rate of the operating current data, and determine the moment when the time change rate first exceeds the preset sudden change threshold as the excitation trigger moment; The difference between the first terminal voltage data at the first sampling time and the second terminal voltage data at the second sampling time in the terminal voltage data is calculated to obtain the transient voltage response value. Both the first sampling time and the second sampling time are adjacent to the excitation triggering time, and the first sampling time is before the excitation triggering time. Calculate the transient current surge amplitude between the first sampling time and the second sampling time based on the operating current data; The ratio of the transient voltage response value to the transient current change amplitude is determined as the ohmic internal resistance value; Obtain the continuous terminal voltage sampling sequence within a preset relaxation time window after the excitation triggering time; A polarization voltage response sequence is generated based on the continuous terminal voltage sampling sequence and the equivalent circuit model, and the polarization impedance value is calculated based on the polarization voltage response sequence.

[0007] Optionally, a polarization voltage response sequence is generated based on the continuous terminal voltage sampling sequence and the equivalent circuit model, and the polarization impedance value is calculated based on the polarization voltage response sequence, specifically including: Based on the equivalent circuit model, the open-circuit voltage state value corresponding to the excitation triggering time is determined. The difference between the target terminal voltage value and the non-polarized composite voltage value at each sampling point in the continuous terminal voltage sampling sequence is calculated respectively. The polarized voltage response sequence after removing the static voltage and ohmic voltage drop is obtained. The non-polarized composite voltage value is the sum of the product of the real-time current value corresponding to the sampling point and the ohmic internal resistance value and the open-circuit voltage state value. The voltage change amplitude of the polarization voltage response sequence within the relaxation time window is obtained, and the ratio of the voltage change amplitude to the transient current change amplitude is calculated to determine the polarization impedance value of the single cell.

[0008] Optionally, the polarization safety factor of each individual cell is determined based on the polarization impedance value, specifically including: Based on the polarization impedance values ​​of all the individual cells, the average group polarization impedance of the target battery pack is calculated; Calculate the impedance dispersion ratio of each individual cell relative to the average polarization impedance of the group; When the impedance dispersion deviation ratio is greater than the preset thermal accumulation warning threshold, the single cell is determined to be in a polarization thermal accumulation risk state, and the ratio of the preset thermal accumulation warning threshold to the impedance dispersion deviation ratio is calculated to obtain the polarization safety factor. When the impedance discrepancy ratio is less than or equal to the preset thermal accumulation warning threshold, the single cell is determined to be in a polarization safety state, and the preset reference coefficient is determined as the polarization safety factor.

[0009] Optionally, the equilibrium demand deviation is determined based on the state of charge value, specifically including: The arithmetic mean of the target battery pack is calculated based on the state of charge values ​​of all the individual cells to obtain the group equilibrium benchmark value, and the original difference between the state of charge value of each individual cell and the group equilibrium benchmark value is calculated. The average of the absolute values ​​of the original differences of all the individual cells is calculated to obtain the group discrete characteristic value, which is used to characterize the statistical average discrete amplitude of the state of charge distribution of each individual cell in the target battery pack. The dead zone radius is determined based on the discrete characteristic values ​​of the population, and the dead zone radius is positively correlated with the discrete characteristic values ​​of the population. A dynamic insensitive dead zone is constructed based on the dead zone radius. The dynamic insensitive dead zone includes a positive cutoff boundary value and a negative cutoff boundary value. The absolute values ​​of the positive cutoff boundary value and the negative cutoff boundary value are both the dead zone radius. If the original difference is within the dynamic insensitivity dead zone, it is determined that the individual cell is within the preset equalization allowable error tolerance range, and the equalization requirement deviation of the individual cell is forcibly set to zero; If the original difference is greater than the positive cutoff boundary value, the difference between the original difference and the positive cutoff boundary is determined as the equilibrium demand deviation degree. If the original difference is less than the negative cutoff boundary value, the difference between the original difference and the negative cutoff boundary value is determined as the equilibrium demand deviation.

[0010] Optionally, the basic balance duty cycle of each individual cell is determined based on the balance demand deviation, specifically including: If the balance demand deviation is zero, the basic balance duty cycle is set directly to zero to keep the balance circuit in the off state. If the equilibrium demand deviation is not zero, then the absolute value of the equilibrium demand deviation is determined as the net equilibrium demand magnitude. The smaller of the ratio of the net balance demand amplitude to the preset full load reference value and the value in Value 1 is determined as the normalized demand coefficient. The preset full load reference value is used to characterize the minimum deviation required for the balance circuit to enter the full load working state. Based on the normalized demand coefficient, affine mapping calculation is performed within the interval formed by the preset effective duty cycle threshold and the preset rated full-load duty cycle to obtain the basic equalization duty cycle. The effective duty cycle threshold is the minimum conduction ratio required for the equalization circuit to maintain effective current output.

[0011] Optionally, the basic equalization duty cycle is corrected based on the polarization safety factor and the aging weighting factor to generate the target pulse width modulation duty cycle, specifically including: Multiplying the polarization safety factor by the aging weighting factor yields the comprehensive state correction factor, which is used to characterize the current carrying capacity weight of the single cell under the constraints of current polarization thermal accumulation risk and aging degree. The theoretical corrected duty cycle is obtained by multiplying the basic equilibrium duty cycle by the comprehensive state correction factor. If the theoretically corrected duty cycle is less than the preset effective duty cycle threshold, the target pulse width modulation duty cycle will be forcibly set to zero; If the theoretically corrected duty cycle is greater than or equal to the preset effective duty cycle threshold, the smaller value between the theoretically corrected duty cycle and the preset hardware safety limit duty cycle is determined as the target pulse width modulation duty cycle.

[0012] Secondly, embodiments of this application provide a battery pack balancing control system based on multidimensional state mapping. The battery pack balancing control system based on multidimensional state mapping includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the battery pack balancing control system based on multidimensional state mapping to perform the method described in the first aspect and any possible implementation thereof.

[0013] Thirdly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a battery pack balancing control system based on multidimensional state mapping, cause the battery pack balancing control system based on multidimensional state mapping to perform the method described in the first aspect and any possible implementation thereof.

[0014] Fourthly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a battery pack balancing control system based on multidimensional state mapping, cause the battery pack balancing control system based on multidimensional state mapping to perform the method described in the first aspect and any possible implementation thereof.

[0015] In summary, one or more technical solutions provided in this application have at least the following technical effects or advantages: 1. By constructing an equivalent circuit model for each individual battery cell and combining real-time operating data such as terminal voltage and operating current, the ohmic internal resistance and polarization impedance values ​​are calculated. Furthermore, the health status is assessed by fusing the ohmic internal resistance growth rate with temperature data, quantifying the aging weighting coefficient to accurately reflect the battery's energy regulation capability boundary during long-term use. Simultaneously, a polarization safety factor is calculated using the polarization impedance value to assess the potential polarization overpotential risk under a given equalization current. Combined with the basic equalization duty cycle formed by voltage or state-of-charge difference, it is further corrected using the aging weighting coefficient and polarization safety factor to generate a target pulse width modulation duty cycle, thereby achieving adaptive adjustment of the equalization current intensity. This not only avoids polarization errors and thermal risks caused by applying excessively strong equalization current to aging batteries but also improves the selectivity and safety of equalization adjustment, ensuring stable and effective energy redistribution of the battery pack under aging heterogeneous conditions. This improves the battery pack's equalization control efficiency.

[0016] 2. By introducing a current step excitation identification mechanism and a multi-segment voltage response analysis method based on an equivalent circuit model, high-precision and dynamic extraction of the ohmic internal resistance and polarization impedance values ​​of a single battery cell is achieved, effectively overcoming the parameter estimation deviation problem caused by the inability to distinguish between static voltage drop and dynamic polarization effect in existing technologies. Specifically, the system first identifies current step excitation events by monitoring the time change rate of the operating current, and calculates the ratio of the voltage jump to the current jump before and after excitation, using the excitation trigger moment as a reference, to accurately obtain the ohmic internal resistance value of the battery. Furthermore, within a preset relaxation time window after excitation, continuous voltage response data is collected, and the open-circuit voltage state is derived by combining the equivalent circuit model. After subtracting the static voltage and ohmic voltage drop, the true polarization voltage response sequence is obtained, and the polarization impedance value is calculated by the ratio of the change amplitude of this sequence to the current jump amplitude. This method can not only dynamically reflect the electrochemical response characteristics of the battery during actual operation, but also effectively identify the nonlinear impedance changes caused by aging or enhanced polarization. It provides accurate basic parameter support for the construction of polarization safety factor and aging weighting factor in subsequent equalization strategies, thereby significantly improving the accuracy, safety and adaptability of battery pack equalization control.

[0017] 3. By introducing a polarization safety factor calculation mechanism based on polarization impedance distribution characteristics, the system achieves early identification and dynamic constraint control of local thermal accumulation risks caused by differences in polarization characteristics within the battery pack, effectively compensating for the lack of risk assessment for the equalization of individual cells with abnormal polarization in existing technologies. Specifically, the system first calculates the average polarization impedance of the group based on the polarization impedance values ​​of all individual cells, serving as a reference for the overall polarization level. Then, it quantifies the degree of polarization characteristic deviation by calculating the impedance dispersion ratio of each individual cell relative to this average value. If the deviation ratio exceeds a preset thermal accumulation warning threshold, the battery is determined to be in a state of polarization thermal accumulation risk, and the polarization safety factor is calculated by the ratio of the deviation ratio to the threshold, which is used to dynamically limit the equalization current intensity it can withstand during equalization control. If the deviation ratio is within the safe range, a preset benchmark coefficient is directly assigned as its polarization safety factor. This method, through quantitative modeling and hierarchical weighted control of polarization risk, not only improves the system's ability to identify high-risk aging batteries, but also effectively prevents local heating and side reaction risks caused by high polarization impedance during the equalization process, thereby enhancing the safety, precision, and adaptive adjustment capability of the entire battery pack equalization control strategy. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a battery pack equalization control method based on multidimensional state mapping in an embodiment of this application. Figure 2 This is a schematic diagram of the process for calculating the ohmic internal resistance and polarization impedance values ​​in the embodiments of this application; Figure 3 This is a schematic diagram of a battery pack balancing control system based on multidimensional state mapping provided in an embodiment of this application.

[0019] Explanation of reference numerals in the attached drawings: 301, Central Processing Unit; 302, Read-Only Memory; 303, Random Access Memory; 304, Bus; 305, Input / Output Interface; 306, Input Section; 307, Output Section; 308, Storage Section; 309, Communication Section; 310, Driver; 311, Removable Media. Detailed Implementation

[0020] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0021] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0022] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0023] Figure 1 This is a flowchart illustrating a battery pack equalization control method based on multidimensional state mapping in an embodiment of this application.

[0024] Please see Figure 1 This application provides a battery pack equalization control method based on multi-dimensional state mapping, the method comprising: S101. Obtain the terminal voltage data, operating current data, and surface temperature data of each individual cell in the target battery pack.

[0025] Step S101 is the foundation for subsequent construction of equivalent circuit model, calculation of battery health state parameters and key indicators of equalization control. It can provide the original input information for the entire equalization control strategy and is also a prerequisite for accurate identification of battery state.

[0026] In practice, a high-precision data acquisition module is deployed in the Battery Management System (BMS). Voltage data is acquired through voltage sampling channels connected to the positive and negative terminals of each individual cell. The sampling circuit typically includes a voltage divider, a voltage follower, and an analog-to-digital converter to achieve real-time and stable measurement of the terminal voltage. Operating current data is obtained by connecting a Hall current sensor or a shunt resistor in series in the battery pack circuit. The Hall current sensor, based on the Hall effect principle, outputs a corresponding voltage signal after detecting the magnetic field generated by the current. The actual operating current value of the battery at the current moment can be calculated. Surface temperature data is acquired by attaching a thermistor (such as an NTC thermistor) to the surface of each individual cell. The resistance of this type of sensor changes with temperature; by measuring the voltage across it and combining this with a known resistance-temperature characteristic curve, the battery surface temperature can be calculated.

[0027] The data acquisition process is triggered at preset time intervals or when key operating conditions change, ensuring that the acquired data has time consistency and dynamic response capabilities. All acquired data undergoes time synchronization and outlier filtering through the BMS's internal data processing module, guaranteeing the accuracy and stability of subsequent modeling and calculations.

[0028] By jointly acquiring three types of data—terminal voltage, operating current, and surface temperature—not only can multi-dimensional modeling of the working state of a single cell be achieved, but it can also provide a reliable basis for the dynamic estimation of parameters such as ohmic internal resistance, polarization impedance, impedance growth rate, and health status, thereby effectively improving the ability of the equalization strategy to perceive the true state of the battery.

[0029] S102. Construct an equivalent circuit model for each of the individual cells. Based on the terminal voltage data, the operating current data, and the equivalent circuit model, calculate the ohmic internal resistance and polarization impedance of each individual cell.

[0030] In step S102, to achieve quantitative modeling and identification of the dynamic electrochemical characteristics inside a single cell, an equivalent circuit model of each cell needs to be constructed. Combined with the collected terminal voltage and operating current data, the ohmic internal resistance and polarization impedance values ​​are further calculated. This calculation process not only characterizes the transient response features of the battery under external excitation but also reflects the internal transport characteristics and polarization behavior of the battery under actual operating conditions, providing a crucial basis for subsequent health status assessment and equalization control.

[0031] In constructing the equivalent circuit model for each individual battery cell, it's crucial to understand that the equivalent circuit model simulates the battery's voltage response and internal dynamic behavior under different operating conditions through combinations of circuit components. This model not only reflects the battery's steady-state performance (e.g., internal resistance) but also describes its dynamic characteristics (e.g., polarization effects and slow recovery processes), serving as the foundation for subsequent parameter identification, state estimation, and control strategy formulation. Commonly used model types include single-RC or multi-RC network models based on the Thevenin structure. The Thevenin model uses a voltage source in series with an ohmic internal resistance, and then connects one or more RC branches in parallel to represent polarization behavior, offering advantages such as simple structure, high computational efficiency, and strong engineering adaptability.

[0032] In practical implementation, the system first collects terminal voltage and current data for each individual battery under multiple operating conditions, including voltage changes and step responses during dynamic charging and discharging. Then, parameter identification algorithms are used to solve for the component parameters in the equivalent circuit model. Common methods include least squares, recursive least squares, or genetic algorithms. Regarding model structure selection, a first-order RC model is preferred, which involves connecting a capacitor and resistor in series in parallel with a voltage source and an ohmic resistor to simulate the polarization process. The voltage source simulates the battery's open-circuit voltage, the ohmic resistor reflects the linear losses of electron and ion flow, and the RC branch simulates the hysteresis characteristics of electrochemical reaction kinetics. The initial model structure can be preset, and the parameters are obtained by fitting historical data and updated in real time during actual operation to reflect changes in battery state. It should be noted that although this embodiment uses a first-order RC equivalent circuit model as an example, this does not constitute a limitation on the scope of protection of this application. In practical applications, to adapt to BMS with different chemical systems (such as lithium iron phosphate and ternary lithium) or different computing power platforms, the equivalent circuit model can also adopt any of the following models: Second-order or multi-order RC models: By connecting multiple RC parallel networks in series, the multi-timescale characteristics of electrochemical polarization inside the battery (such as the separation of electrochemical polarization and concentration polarization) can be simulated more accurately. PNGV (Partnership for a New Generation of Vehicles) model: Based on the RC model, a capacitor is added to describe the characteristics of open-circuit voltage as a function of load charge. Thevenin model: the classic Thevenin equivalent circuit structure; Data-driven models: Black-box models built using neural networks (such as LSTM, RNN) or support vector machines (SVM) directly map the current internal resistance and polarization state parameters to the input voltage and current history sequences. Any model or method capable of separating the ohmic internal resistance and polarization impedance characteristics based on terminal voltage and operating current is covered within the scope of this application. The aforementioned model is not only applicable to state estimation under static conditions but also provides accurate voltage prediction capabilities when current changes rapidly, thus providing theoretical support for the extraction of ohmic internal resistance and polarization impedance. For example, if the fitting results of voltage-current data collected during the charging and discharging process of a single cell from 0.5A to 5A show that its optimal fitting model is a first-order RC model, where the voltage source value is 3.8V, the ohmic resistance is 5.6mΩ, and the resistance and capacitance in the polarization branch are 12mΩ and 2300μF, this equivalent circuit model will serve as the basic model for subsequent state mapping and equalization control of this single cell, participating in key processes such as ohmic voltage drop stripping, polarization voltage extraction, and state of charge estimation. By constructing this model, complex electrochemical processes can be transformed into analytically calculable circuit behaviors, thereby improving the system's ability to understand and control the battery state.

[0033] Figure 2 This is a flowchart illustrating the calculation of the ohmic internal resistance and polarization impedance values ​​in the embodiments of this application. The following is a summary of the process. Figure 2 The following provides a detailed explanation of how to calculate the ohmic internal resistance and polarization impedance of each individual cell in step S102 based on the terminal voltage data, the operating current data, and the equivalent circuit model.

[0034] S201. Monitor the time change rate of the operating current data, and determine the moment when the time change rate first exceeds the preset mutation threshold as the excitation trigger moment.

[0035] To accurately identify the moment when a single cell is in a current step excitation state and use this as a time reference for subsequent voltage response feature extraction, it is necessary to monitor the time change rate of the operating current data in real time and combine it with a preset abrupt change threshold for judgment. Current step excitation is a commonly used dynamic testing method in the field of battery modeling. Its principle is to identify its internal parameters, such as ohmic internal resistance and polarization impedance, by using the response characteristics of the battery terminal voltage to the abrupt change when a significant current change occurs during battery operation. Therefore, the current change rate, as the core indicator for judging whether a step excitation has occurred, is of crucial significance for the subsequent identification of electrochemical parameters.

[0036] In practice, the battery management system (BMS) uses a built-in high-frequency sampling current sensor to collect operating current data in real time. This data is then differentially processed at fixed time intervals to form a current-time rate of change sequence. A pre-set threshold for abrupt current changes is established within the system, typically based on the battery pack's rated capacity or historical operating data. For example, it could be set to a rate of change range of 5-10 times the current average operating current. When the current rate of change at the current time point is detected to exceed this threshold, it indicates that the battery is undergoing a significant switching of operating states or a sudden load change, satisfying the step excitation condition. Upon confirmation, the system records this moment as the excitation trigger moment and locks it as a reference time point for subsequent voltage response extraction.

[0037] S202. Calculate the difference between the first terminal voltage data at the first sampling time and the second terminal voltage data at the second sampling time in the terminal voltage data to obtain the transient voltage response value. Both the first sampling time and the second sampling time are adjacent to the excitation triggering time, and the first sampling time is before the excitation triggering time.

[0038] Because the battery's terminal voltage experiences a momentary jump when driven by a sudden current change, and this jump is mainly caused by the ohmic internal resistance, a typical transient voltage response value can be obtained by selecting two adjacent sampling times before and after the excitation trigger moment and calculating the difference between the corresponding terminal voltage data. This transient voltage response value, as a direct reflection of the ohmic voltage drop, forms the basis for subsequent calculations of the ohmic internal resistance value.

[0039] In practice, the process begins by using the previously determined excitation trigger time to search forward one sampling period and simultaneously searching backward one sampling period for the second sampling time. The first sampling time is the last voltage sampling time before the current surge, and the second sampling time is the first voltage sampling time after the current surge; the two are typically separated by one data sampling period. The terminal voltage values ​​corresponding to these two sampling times are then extracted from the cached terminal voltage data sequence and denoted as the first terminal voltage data and the second terminal voltage data. By subtracting these two voltage values, the transient voltage response value reflects the voltage jump amplitude generated inside the battery under the influence of the current surge. This value can be directly used to compare with the current surge amplitude during the same period to determine the ohmic internal resistance.

[0040] This processing method not only enables dynamic estimation of battery internal resistance during actual operation but also avoids the complex procedures of introducing external excitation sources or interrupting system operation, demonstrating excellent real-time performance and engineering feasibility. By accurately extracting transient voltage response values, the polarization component and static open-circuit voltage component contained in the battery terminal voltage can be effectively separated, making the obtained ohmic internal resistance value more representative and providing accurate basic parameters for subsequent polarization impedance extraction and state identification.

[0041] S203. Calculate the transient current change amplitude between the first sampling time and the second sampling time based on the operating current data.

[0042] To accurately calculate the current change amplitude experienced by a single cell during a current step excitation process, it is necessary to extract the current values ​​corresponding to the first and second sampling times based on the operating current data, and calculate the transient current surge amplitude between these two times. This surge amplitude is a key parameter describing the current change during the transition from one stable operating state to another, and is directly used to calculate the ratio with the transient voltage response value to determine the ohmic internal resistance. Therefore, the accurate identification of the transient current surge amplitude not only affects the ohmic internal resistance calculation result but also determines the benchmark accuracy of subsequent polarization impedance extraction.

[0043] In practice, after the current step excitation is identified, the system first locks the first and second sampling times, i.e., two consecutive sampling points before and after the excitation trigger moment. Then, the current values ​​at these two times are extracted from the real-time operating current data sequence and recorded as the first-time current value and the second-time current value, respectively. The amplitude of the change, i.e., the transient current surge amplitude, is obtained by performing a difference calculation on these two values. Typically, this value is positive or negative, depending on the direction of the current change, but its absolute value is usually taken in subsequent calculations of ohmic resistance and polarization impedance.

[0044] The above process is based on a high-sampling-rate current sensor in the battery management system, with common sampling frequencies ranging from 100Hz to 1kHz, ensuring accurate capture of sudden changes. Since this step only involves calculating current values ​​at two known moments, the computational load is small and the response is fast, meeting the real-time requirements of embedded systems. Furthermore, the amplitude of this current surge is engineering-interpretable, quantifying the intensity of the current excitation event and providing auxiliary reference for subsequent assessment of the effectiveness of the voltage response.

[0045] S204. The ratio of the transient voltage response value to the transient current change amplitude is determined as the ohmic internal resistance value.

[0046] The ohmic internal resistance is an important parameter reflecting the resistive losses in the internal conductive paths of a battery (including electrodes, current collectors, and electrolytes). Its changes directly affect the battery's charge and discharge efficiency, thermal characteristics, and aging condition. During actual battery operation, when the current changes abruptly, the terminal voltage will immediately experience a rapid change. This change mainly originates from the ohmic voltage drop. According to Ohm's law, this voltage drop is proportional to the change in current, and the proportionality constant is the ohmic internal resistance. Therefore, the internal resistance can be calculated by comparing the two.

[0047] In practical implementation, this process first uses the transient voltage response value obtained in step S202 as the amplitude of the terminal voltage jump caused by the current surge, while simultaneously calling the transient current surge amplitude calculated in step S203. The system substitutes both into Ohm's law formula: R = ΔU / ΔI, where R is the internal resistance to be calculated, ΔU is the transient voltage response value, and ΔI is the transient current surge amplitude. The ratio calculation is completed by the digital signal processing module, typically using floating-point arithmetic to improve accuracy, and the calculation result is written to the battery parameter database in real time for subsequent health assessment and balancing strategy use.

[0048] The advantage of this method lies in its reliance on the natural current excitation under actual operating conditions, eliminating the need for additional excitation signals. Furthermore, it leverages high-precision voltage and current sampling data to achieve online identification of the battery's internal resistance, exhibiting excellent real-time performance and engineering adaptability. Changes in internal resistance can also serve as a key indicator of battery aging, helping to determine the presence of potential safety hazards such as localized thermal runaway or interface degradation.

[0049] S205. Obtain the continuous terminal voltage sampling sequence within the preset relaxation time window after the excitation trigger time.

[0050] In step S205, to accurately identify the battery polarization impedance value, a complete voltage recovery process needs to be acquired after the excitation trigger moment to construct the polarization voltage response sequence. Since polarization impedance is a delayed response characteristic of the battery's internal electrochemical process to current changes, it is usually manifested as a dynamic process in which the voltage slowly stabilizes after a sudden current change. Therefore, it is necessary to continuously collect terminal voltage data within a certain time range after the current step excitation occurs to form a complete relaxation voltage change curve, which serves as the basis for subsequent extraction of polarization response characteristics.

[0051] In practice, once the system identifies the excitation trigger moment, timing begins, and a fixed-length time window, called the relaxation time window, is set after that moment. The purpose of setting the relaxation time window is to capture the dynamic changes in the battery's terminal voltage as it gradually recovers from a sudden current jump to a steady state. This process primarily reflects the response behavior of the battery's internal polarization phenomenon. Since the polarization effect originates from electrochemical reaction kinetics and ion migration processes and has a certain time lag, it is necessary to continuously sample the terminal voltage at high frequency for a period after the excitation trigger moment to form a complete voltage relaxation curve. The length of this time window is usually set empirically based on the time constant of the polarization branch in the equivalent circuit model, the battery type, and the actual response speed, ensuring that the collected data effectively covers the polarization voltage change process, thus providing sufficient data support for the subsequent extraction of polarization impedance values. Within this time window, the system continuously collects terminal voltage data at a fixed sampling frequency (such as 200Hz or higher) through the high-precision voltage sampling channel in the battery management controller, and stores the voltage values ​​of all sampling points in the buffer in chronological order to form a complete continuous terminal voltage sampling sequence, which serves as the raw input data for polarization response analysis.

[0052] This sampling sequence not only records the rapid voltage change of the battery in the initial stage after a sudden current change, but also covers the entire process of the voltage gradually approaching a steady state in the middle and later stages. Therefore, it can reflect the true dynamic characteristics of the polarization process and help to isolate polarization behavior that is independent of ohmic voltage drop. By subsequently fitting the polarization branch response in the equivalent circuit model, the magnitude of the polarization impedance can be effectively estimated, providing a key basis for battery aging status evaluation and equalization control.

[0053] S206. Generate a polarization voltage response sequence based on the continuous terminal voltage sampling sequence and the equivalent circuit model, and calculate the polarization impedance value based on the polarization voltage response sequence.

[0054] In step S206, to further extract the gradually changing voltage response characteristics exhibited by a single cell after a sudden current change, it is necessary to generate a polarization voltage response sequence that characterizes polarization behavior based on the acquired continuous terminal voltage sampling sequence and the constructed equivalent circuit model. This polarization voltage response sequence reflects the dynamic changes of the internal electrochemical process of the battery during the relaxation phase and is the basis for identifying the polarization impedance value. Therefore, in this step, it is necessary to further remove the static open-circuit voltage component and the ohmic voltage drop component from the voltage sequence to obtain the voltage change process purely caused by the polarization effect. Based on this, the polarization voltage response amplitude is extracted and compared with the transient current change amplitude to finally determine the polarization impedance value. Specifically, this may include the following steps: determining the open-circuit voltage state value corresponding to the excitation triggering time based on the equivalent circuit model; calculating the difference between the target terminal voltage value and the non-polarized composite voltage value at each sampling point in the continuous terminal voltage sampling sequence to obtain the polarized voltage response sequence after removing the static voltage and ohmic voltage drop; the non-polarized composite voltage value is the sum of the product of the real-time current value corresponding to the sampling point and the ohmic internal resistance value and the open-circuit voltage state value; obtaining the voltage change amplitude of the polarized voltage response sequence within the relaxation time window; calculating the ratio of the voltage change amplitude to the transient current change amplitude to determine the polarization impedance value of the single cell.

[0055] The open-circuit voltage state value refers to the stable voltage value of a battery under conditions of no current flow, reflecting the intrinsic potential of the battery under its current electrochemical state (such as remaining capacity and concentration). Since the battery is always under load during actual operation, the open-circuit voltage cannot be directly measured and needs to be estimated using an equivalent circuit model. The equivalent circuit model typically includes a voltage source, an ohmic resistance branch, and a first- or multi-order RC polarization branch, which can simulate the overall voltage response characteristics of the battery under dynamic load. During implementation, after identifying the excitation trigger moment, the system performs a short-time fitting of the terminal voltage and current data prior to that moment, and combines this with the identified ohmic internal resistance value to eliminate the influence of transient voltage drop, thereby calculating the theoretical open-circuit voltage state value before the current effect at the excitation trigger moment causes polarization. This value serves as a static reference for subsequent polarization voltage response extraction, helping to separate the slowly varying voltage components caused by polarization effects, laying the foundation for accurate identification of polarization impedance.

[0056] After obtaining the open-circuit voltage state value, each voltage sampling point within the relaxation time window needs to be processed. Since the actual measured terminal voltage value contains multiple superimposed components, including the open-circuit voltage, ohmic voltage drop, and polarization voltage response, the non-polarization component needs to be separated. Specifically, based on the real-time operating current value corresponding to each sampling point, the identified ohmic internal resistance value is multiplied to obtain the ohmic voltage drop at that moment. This is then added to the open-circuit voltage state value to obtain the non-polarization composite voltage value for that sampling point, i.e., the theoretical voltage level without considering polarization effects. Subsequently, the difference between this non-polarization composite voltage value and the actual acquired target terminal voltage value is taken as the polarization voltage response value for that sampling point. Applying the above operations to the entire continuous terminal voltage sampling sequence allows the construction of a complete polarization voltage response sequence. This sequence accurately reflects the voltage change caused by the polarization process after a current step excitation of the battery and is the core data structure upon which polarization impedance calculation relies.

[0057] After successfully extracting the polarization voltage response sequence, the system needs to further extract numerical features from this sequence for impedance calculation. Since polarization impedance is essentially the ratio of voltage change to current change during polarization, the amplitude of the voltage change within the entire relaxation time window—that is, the maximum voltage difference from the start to the end of the polarization response—is used as an indicator of the polarization response amplitude. This voltage change amplitude typically exhibits a negative, gradual recovery to a steady state; the larger the value, the more significant the battery polarization phenomenon. Subsequently, the ratio of this voltage change amplitude to the transient current jump amplitude extracted in the previous step is calculated, and the result is the polarization impedance value. Because the transient current jump amplitude has been accurately extracted at the time of the current step, its ratio to the polarization voltage response amplitude effectively reflects the voltage response capability of the internal polarization branch of the battery under unit current excitation, thus characterizing the magnitude of the polarization impedance.

[0058] For example, in a single-cell current surge event, the system identifies an open-circuit voltage state of 3.795V at the excitation trigger moment, a real-time operating current of 2.4A at a sampling point within the relaxation time window, and an internal resistance of 6mΩ. Therefore, the non-polarized composite voltage at this point is 3.795V + 2.4A × 0.006Ω = 3.8094V, and the actual sampling terminal voltage is 3.785V. Thus, the polarization voltage response at this point is 3.785V − 3.8094V = −0.0244V. If the maximum voltage change amplitude of the entire polarization voltage response sequence is 0.036V, and the transient current surge amplitude is 4.5A, then the polarization impedance is 0.036V / 4.5A = 0.008Ω, or 8mΩ. This parameter will be used in subsequent polarization safety factor calculations and equalization control strategy corrections.

[0059] S103. Calculate the impedance growth rate of the ohmic internal resistance value relative to the preset nominal internal resistance value.

[0060] In step S103, to quantitatively assess the degree of health degradation of a single battery cell, it is necessary to calculate the percentage change in its internal resistance relative to the nominal internal resistance value set at the initial factory state, based on the ohmic internal resistance value identified in the previous steps. This percentage change is known as the impedance growth rate. The impedance growth rate, as a crucial indicator of battery lifespan changes, reflects resistive degradation caused by factors such as damage to the internal electron conduction path and ion transport interface, material aging, or current collector corrosion. Calculating the impedance growth rate provides a basis for subsequent weighting of health status values, determination of aging weighting coefficients, and dynamic adjustment of balancing strategies.

[0061] In the specific implementation process, the nominal internal resistance value is first extracted from the battery's factory technical data or early-life experimental tests. This nominal internal resistance value refers to the benchmark ohmic internal resistance value measured by the new battery under standard temperature and rated charging conditions, which is representative and comparable. Then, the currently identified ohmic internal resistance value is compared with the nominal internal resistance value, and the impedance growth rate is calculated according to the following formula: Impedance growth rate = (Ohmic internal resistance value - Nominal internal resistance value) ÷ Nominal internal resistance value. This calculation process is usually automatically completed by the state estimation module in the battery management system. Once a valid ohmic internal resistance value is identified, a preset nominal value can be called for real-time calculation. The impedance growth rate result is expressed as a percentage; the higher the value, the more severe the internal impedance degradation of the battery, and the higher the energy consumption, heat generation risk, and reduced output capacity.

[0062] For example, if a single battery cell currently has an identified internal resistance of 7.2 milliohms and a nominal internal resistance of 5.0 milliohms, then the impedance growth rate is: (7.2 − 5.0) ÷ 5.0 = 0.44, or 44%. This result indicates that the battery's conduction loss has increased by 44% compared to its initial state. Subsequent calculations will be based on this increase in its state of health, and a lower current regulation weight will be assigned to it in the equalization control to prevent further accelerated aging or the risk of localized overheating. This method enables refined identification and dynamic adjustment control of the battery state, improving the safety and consistency of the entire battery pack operation.

[0063] S104. Based on the impedance growth rate and the surface temperature data, determine the health status value of each individual cell, and determine the aging weighting coefficient based on the health status value of each individual cell.

[0064] In step S104, to achieve refined identification and dynamic equalization control of the aging degree of each individual cell in the battery pack, it is necessary to determine the health state value of each individual cell based on impedance growth rate and surface temperature data, and further set an aging weighting coefficient according to this health state value. The health state value is a standardized indicator used to quantitatively describe the current performance degradation degree of the battery, usually represented by a continuous value between 0 and 1, where 1 represents good performance and 0 represents complete failure. The aging weighting coefficient is an adjustment factor derived from the health state value, mainly used to assign different equalization weights to batteries with different aging degrees in the equalization control strategy to prevent secondary aging or safety hazards caused by over-equalization.

[0065] In the specific implementation process, the impedance growth rate calculated in the previous steps is first used as the main indicator to measure the degradation of the internal conductive path. A higher impedance growth rate indicates a more severe limitation on the battery's electron conduction efficiency and ion transport path, intensified internal polarization, and increased risk of heat generation. Simultaneously, surface temperature data is introduced as an auxiliary criterion because temperature not only reflects the battery's current thermal state but also indirectly reveals its thermal management capabilities and internal chemical stability during long-term cycling. By jointly mapping the impedance growth rate and surface temperature data, the accuracy and robustness of the health status assessment can be effectively improved, especially in high- or low-temperature operating environments, helping to avoid misjudgments based on a single indicator.

[0066] In the modeling implementation, to establish the mapping relationship between health state values, impedance growth rate, and surface temperature, a multidimensional interpolation table model can be built using two-dimensional linear piecewise interpolation. Specifically, battery performance data is first collected through numerous experiments under different combinations of impedance growth rates (e.g., 0% to 100%) and surface temperatures (e.g., 20℃ to 60℃). Based on these sample points, corresponding health state values ​​are manually set or determined through degradation model analysis, forming a two-dimensional feature-response data matrix. Subsequently, a two-dimensional interpolation table is constructed with impedance growth rate as the horizontal axis, surface temperature as the vertical axis, and health state value as the interpolation result. During operation, the impedance growth rate and temperature of the current individual cell are used as inputs in real time, and the corresponding health state value is obtained through bilinear interpolation. For example, when the impedance growth rate is 40% and the surface temperature is 45℃, the system quickly queries and outputs a health state value of 0.62 through the interpolation table, which serves as the basis for subsequent aging weighted calculations. Furthermore, to quantify the aging weighting coefficients, the system can use the following function model for calculation: ,in, The aging weighting factor is... This is the health status value obtained from the current query. The preset aging inflection point threshold (for example, a value of 0.8 indicates that derating begins when the battery health is below 80%).

[0067] S105. Determine the polarization safety factor for each individual cell based on the polarization impedance value.

[0068] The polarization safety factor is a regulatory factor used to reflect the degree of thermal accumulation risk that a battery may cause due to polarization under current operating conditions, and serves as an important basis for subsequent equalization duty cycle correction. Since a larger polarization resistance value indicates a more significant hysteresis effect in the electrochemical reaction, which easily leads to local heat accumulation, it is necessary to reduce its participation intensity during the equalization process of this type of battery to prevent thermal runaway. Therefore, this step further introduces a criterion mechanism based on the average polarization impedance of the group and the degree of deviation of individual cells. By comparing and analyzing the polarization impedance value of each individual cell with the overall level of the battery pack, it identifies whether the battery is in a state of polarization thermal accumulation risk, and calculates or sets its corresponding polarization safety factor accordingly. Specifically, it may include the following steps: calculating the average polarization impedance of the target battery pack based on the polarization impedance values ​​of all the individual cells; calculating the impedance dispersion deviation ratio of the polarization impedance value of each individual cell relative to the average polarization impedance of the group; when the impedance dispersion deviation ratio is greater than a preset thermal accumulation warning threshold, determining that the individual cell is in a state of polarization thermal accumulation risk, and calculating the ratio of the preset thermal accumulation warning threshold to the impedance dispersion deviation ratio to obtain the polarization safety factor; when the impedance dispersion deviation ratio is less than or equal to the preset thermal accumulation warning threshold, determining that the individual cell is in a polarization safe state, and setting a preset benchmark coefficient as the polarization safety factor.

[0069] During implementation, to identify individual cells in the battery pack that may experience localized thermal buildup due to polarization effects, and thus dynamically limit their equalization current, it is necessary to further calculate the average polarization impedance of the entire target battery pack based on the polarization impedance values ​​currently identified for each individual cell. Polarization impedance is a crucial parameter for evaluating the hysteresis behavior of the internal electrochemical reactions of a battery. It is typically extracted from the voltage response following a current step excitation in a previous step, reflecting the additional voltage loss caused by polarization. Since the polarization impedance values ​​of individual cells within the battery pack may vary significantly due to factors such as aging, thermal management efficiency, and operating conditions, it is necessary to calculate a group average as a reference benchmark for the overall polarization level of the system. In practice, the system sums the polarization impedance values ​​of each individual cell in the target battery pack and divides the sum by the total number of cells to obtain the average group polarization impedance at the current moment, which is used for subsequent differential analysis.

[0070] After obtaining the average polarization impedance of the group, to further identify potential polarization-abnormal cells, it is necessary to calculate the dispersion ratio of the polarization impedance value of each individual cell relative to the group average. This deviation ratio is obtained by subtracting the group average from the individual cell's polarization impedance value and then dividing by the group average. It is a standardized relative deviation index that reflects the degree of deviation of an individual cell's polarization characteristics from the average level of the battery pack. A higher deviation ratio indicates that the polarization behavior of the cell is more prominent in its current state, which may be due to passivation of internal active materials, electrolyte imbalance, or deterioration of electrode pore structure, leading to the accumulation of more polarization heat during current regulation and posing a potential safety hazard. By introducing this deviation ratio, early identification of abnormal cells can be achieved, laying the foundation for differentiated adjustment of subsequent control strategies.

[0071] After calculating the deviation ratio, the system compares it with a preset thermal accumulation warning threshold to determine whether a single cell is in a polarization thermal accumulation risk state. This thermal accumulation warning threshold is a critical reference value set through extensive experiments or simulation modeling, typically reflecting the maximum polarization deviation the battery can withstand under current thermal management capabilities. When the deviation ratio exceeds this threshold, the battery is considered to be in a polarization thermal accumulation risk range. To ensure its safe operation, its ability to participate in the equalization process needs to be limited. Specifically, the ratio between the thermal accumulation warning threshold and the deviation ratio is used as a polarization safety factor. This factor is usually less than 1; the smaller the value, the higher the risk. In subsequent equalization strategies, the current regulation intensity allocated to the battery will be significantly reduced, thereby effectively weakening its polarization heat accumulation trend, delaying local aging, and reducing the risk of thermal runaway.

[0072] If the deviation ratio is lower than or equal to the thermal accumulation warning threshold, the system determines that the individual cell is in a polarization safe state. In this case, to avoid unnecessary power reduction and increased control complexity, a default polarization safety reference coefficient can be directly assigned to it. This reference coefficient is usually set to 1, indicating that it can participate in the equalization current regulation at its normal duty cycle without being restricted by polarization risks. This graded regulation mechanism allows the battery pack to maximize equalization efficiency while ensuring safety, and improves overall operational consistency and lifespan utilization.

[0073] For example, if the current polarization impedance of a single cell is 4.8 mΩ, while the average polarization impedance of the entire battery pack is 3.0 mΩ, then its impedance dispersion deviation ratio is (4.8 − 3.0) ÷ 3.0 = 0.6. If the preset thermal accumulation warning threshold is 0.4, then the deviation ratio of this cell has significantly exceeded the standard, and the system classifies it as a high-risk cell. The system further calculates the polarization safety factor as 0.4 ÷ 0.6 ≈ 0.67, indicating that this cell is only allowed to participate in equalization control at 67% of its original duty cycle, effectively reducing the heat generation and aging rate caused by excessively high polarization impedance. Through this method, real-time identification and dynamic response control of polarization risks can be achieved, improving the safety, stability, and refined management capabilities of the battery pack operation.

[0074] S106. Determine the state of charge (SOC) value of each individual cell in the target battery pack based on the preset charge calculation model, and determine the balance requirement deviation based on the SOC value.

[0075] In step S106, to comprehensively evaluate the energy distribution characteristics of each individual cell within the target battery pack and identify whether any cells require charge migration adjustment, it is necessary to calculate the deviation of each individual cell's State of Charge (SOC) relative to the overall level, thereby determining the corresponding balance demand deviation. The State of Charge is a core parameter for measuring the remaining usable capacity of the battery, calculated using a preset SOC calculation model. The core purpose of this model is to dynamically estimate the current percentage of remaining usable capacity of the battery by fusing measurement data with the battery characteristic model, thus providing basic data support for subsequent balance demand judgment. The following section explains its implementation logic from the perspective of model principles and structural composition.

[0076] The charge calculation model uses a fusion model combining the current integral method and voltage correction as its basic framework. In this model, the current integral method is used to capture the net change in charge of the battery during charging and discharging. Its basic principle is to multiply the current per unit time by the time length to obtain the change in charge, and then recursively deduce the SOC value based on this. The specific formula is as follows: ,in, The initial state of charge, This refers to the battery's nominal capacity. This represents the instantaneous current; a positive value indicates discharge, and a negative value indicates charging. This integration process is achieved through high-frequency sampling, enabling continuous tracking of battery charge changes.

[0077] To address the error accumulation issue inherent in the current integration method during long-term operation, a voltage correction mechanism is introduced as an auxiliary calibration path. This mechanism leverages the mapping relationship between open circuit voltage (OCV) and state of charge (SOC) under static conditions. By comparing the real-time measured terminal voltage with the pre-defined OCV-SOC curve in the model, it determines whether the SOC calculated by the integration method has deviated and corrects it accordingly. This OCV-SOC relationship curve is typically obtained through experimental calibration, exhibits monotonicity, and is suitable for estimation and calibration under stable or static current conditions.

[0078] Furthermore, to adapt to the varying characteristics of different types of battery cells, the model also incorporates a temperature compensation mechanism and an adaptive aging parameter module. The temperature compensation mechanism corrects for OCV curve drift and internal resistance changes caused by temperature variations, ensuring the accuracy of SOC estimation under extreme environments such as low and high temperatures. The aging parameter module dynamically adjusts the capacity parameters in the model by tracking the internal resistance growth rate and capacity decay trend. This is to reflect the impact of performance degradation during the battery's lifespan on SOC estimation.

[0079] However, due to factors such as uneven load distribution and inconsistent aging, the SOC distribution of battery packs gradually becomes unbalanced during long-term operation. Therefore, relying solely on the SOC value is insufficient to determine whether balancing operations are needed. To address this, this step introduces a deviation identification method based on a combination of group statistical characteristics and a dynamic dead-zone mechanism. By jointly calculating the mean SOC, discrete amplitude, and individual deviations of all individual cells, a constrained balancing demand model with low sensitivity is constructed. This effectively suppresses ineffective balancing caused by small deviations, improving the energy efficiency and stability of the control strategy. Specifically, this may include the following steps: The arithmetic mean of the target battery pack is calculated based on the state of charge (SOC) values ​​of all individual cells to obtain a group equilibrium benchmark value. The original difference between the SOC value of each individual cell and the group equilibrium benchmark value is then calculated. The average of the absolute values ​​of all the original differences is calculated to obtain a group discrete characteristic value, which characterizes the statistically average discrete amplitude of the SOC distribution of each individual cell within the target battery pack. A dead zone radius is determined based on the group discrete characteristic value, and this dead zone radius is positively correlated with the group discrete characteristic value. A dynamically insensitive dead zone is constructed based on the dead zone radius. The dead zone includes a positive cutoff boundary value and a negative cutoff boundary value, where the absolute values ​​of both the positive and negative cutoff boundary values ​​are the dead zone radius. If the original difference is within the dynamic insensitivity dead zone, the individual cell is determined to be within a preset equalization allowable error tolerance range, and the equalization requirement deviation of the individual cell is forcibly set to zero. If the original difference is greater than the positive cutoff boundary value, the difference between the original difference and the positive cutoff boundary value is determined as the equalization requirement deviation. If the original difference is less than the negative cutoff boundary value, the difference between the original difference and the negative cutoff boundary value is determined as the equalization requirement deviation.

[0080] In practical implementation, to accurately identify the necessity for each individual cell in the target battery pack to participate in equalization adjustment, it is necessary to rely on their State of Charge (SOC) distribution analysis and construct a standardized index to measure the degree of equalization demand—the equalization demand deviation. State of Charge is an important parameter for measuring the remaining capacity of a battery, usually expressed as a percentage of the battery's current usable capacity relative to its rated capacity. Although SOC itself can be used to determine the battery's capacity, due to manufacturing differences, varying aging rates, and uneven thermal environments among the individual cells in the battery pack, their SOC distribution often exhibits a certain degree of dispersion. Therefore, relying solely on the original SOC value cannot effectively guide the refined execution of the equalization strategy.

[0081] To address this issue, in this step, the system first performs an arithmetic average of the State of Charge (SOC) of all individual cells to obtain the current group equilibrium benchmark value of the battery pack. This value serves as the ideal equilibrium center for the entire battery pack. Subsequently, the difference between the SOC of each individual cell and this benchmark value is calculated, yielding the original difference for each cell. This difference directly reflects the degree of deviation from the overall state of the battery pack. To further quantify the overall dispersion of the SOC distribution within the battery pack, the system averages the absolute values ​​of all original differences to obtain a statistically significant index called the group dispersion characteristic value. This value characterizes the average deviation of the current battery pack in energy distribution.

[0082] Based on the aforementioned discrete characteristic values ​​of the battery pack, the system further sets a dead zone radius, which defines an allowable error tolerance range to prevent energy waste and control resource consumption caused by frequent triggering of equalization operations due to small differences. This dead zone radius is positively correlated with the discrete characteristic values ​​of the battery pack; that is, the more dispersed the overall SOC distribution of the battery pack, the larger the dead zone range becomes, in order to maintain the adaptability and robustness of the strategy.

[0083] In the implementation process, the dead zone radius is determined based on the discrete characteristic values ​​of the group. Essentially, this involves mapping the dispersion of the current SOC distribution within the battery pack to the tolerance range in the equalization control strategy using an adaptive function model, thereby dynamically suppressing small SOC deviations. As the control boundary of the dynamically insensitive dead zone, the determination of the dead zone radius must consider not only the fluctuation range of the original difference but also the changing trend of the overall consistency level of the battery pack. Therefore, it is necessary to scale or nonlinearly map the discrete characteristic values ​​of the group. The specific implementation method is as follows: The system first obtains the discrete characteristic value of the group, calculated by the absolute value of the average original difference. This value can be denoted as D_group, and the unit is usually percentage (%), reflecting the average deviation of the SOC of each individual cell from the mean. In order to transform this statistical result into a dead zone threshold that can be used for control judgment, the system presets a dead zone ratio coefficient, denoted as k_deadzone. This coefficient is a dimensionless quantity, and its value is generally in the range of 0.8 to 1.5. The specific value is set according to the system's requirements for equalization sensitivity. For example, in the case of high energy efficiency requirements and large deviation tolerance, a larger value can be set to avoid frequent equalization.

[0084] Subsequently, the dead zone radius R_deadzone is calculated according to the following linear mapping function: R_deadzone = k_deadzone × D_group, where R_deadzone is the final determined dead zone radius, which determines the upper and lower limits of the dynamically insensitive dead zone. This mapping relationship has good physical interpretability: when the battery pack's SOC distribution is relatively concentrated (i.e., D_group is small), the dead zone is set narrower, and the equalization strategy is more aggressive; while when the SOC distribution has shown strong dispersion (i.e., D_group is large), the system automatically relaxes the judgment criteria, expands the dead zone radius, reduces ineffective equalization, and improves energy utilization efficiency.

[0085] In some advanced applications, nonlinear function mapping models, such as exponential functions or piecewise linear functions, can be used to improve adaptability to extreme cases. For example, R_deadzone = a × (1 - e^(−b × D_group)), where a and b are empirical adjustment parameters used to control the rate and upper limit of dead zone growth. When D_group approaches 0, the dead zone radius approaches 0; when D_group increases, the dead zone radius grows nonlinearly, reflecting an improved tolerance to highly discrete states. For example, if the currently calculated population discrete characteristic value is 1.6%, and the preset dead zone ratio coefficient is 1.2, then the dead zone radius is: R_deadzone = 1.2 × 1.6% = 1.92%. Therefore, the positive cutoff boundary of the dynamically insensitive dead zone is determined to be +1.92%, and the negative cutoff boundary is -1.92%. When determining the equalization requirement, any single cell whose SOC deviates within ±1.92% is considered to have met the tolerance requirement and does not need to be equalized, thereby effectively reducing unnecessary energy migration operations and improving the overall operating efficiency and response stability of the system.

[0086] To effectively identify and classify the state-of-charge (SOC) deviations of individual cells in the battery pack, a dynamically insensitive dead zone needs to be constructed based on the previously determined dead zone radius. This dead zone serves as a tolerance mechanism for judging equalization requirements. The insensitive dead zone is a judgment interval that tolerates SOC differences. Its design aims to shield against ineffective equalization behavior caused by small SOC deviations, thereby improving the energy efficiency ratio and system stability of the control strategy. Specifically, the system first constructs a symmetrical interval along the SOC deviation dimension based on the calculated dead zone radius R_deadzone, using this as the boundary to determine the upper and lower limits of the dynamically insensitive dead zone. The positive cutoff boundary value of this interval is set as +R_deadzone, and the negative cutoff boundary value is set as −R_deadzone. Both absolute values ​​are equal, representing the dead zone radius. This symmetrical interval construction ensures a unified tolerance standard for individual cell deviations above and below the group equalization benchmark value, avoiding control bias.

[0087] Subsequently, the system judges the original difference (i.e., the difference between its SOC value and the group equilibrium benchmark value) of each individual cell. If the value of the original difference falls between the positive and negative cutoff boundaries, that is, satisfying −R_deadzone < original difference < +R_deadzone, it indicates that the current state of charge of the cell is within an acceptable range. The system determines that it is within the preset equilibrium allowable error tolerance range and does not need to participate in this round of equilibrium control. Therefore, its corresponding equilibrium demand deviation is forcibly set to zero to avoid resource waste.

[0088] For individual cells whose initial difference exceeds the boundary of the insensitive zone, the system identifies them as target objects with significant energy deviation and calculates the corresponding balance demand deviation based on the magnitude of their deviation from the boundary. Specifically, if the initial difference is greater than the positive cutoff boundary value, i.e., initial difference > +R_deadzone, it indicates that the cell's SOC is significantly higher than the balance benchmark, and energy release is required. The deviation is calculated as: Balance Demand Deviation = Initial Difference - Positive Cutoff Boundary Value.

[0089] Similarly, if the initial difference is less than the negative cutoff boundary value, i.e., the initial difference < −R_deadzone, it indicates that the battery's SOC is significantly lower than the balancing benchmark, requiring energy compensation. The deviation is calculated as: Balancing Demand Deviation = Initial Difference − Negative Cutoff Boundary Value. Through the above difference operation between the initial difference and the boundary, the system not only preserves the directionality (positive or negative) of the initial SOC deviation but also achieves the quantitative extraction of the effective balancing demand. This deviation will be used in subsequent steps to determine the energy release intensity that the battery should bear in the balancing adjustment. Conversely, if the initial difference of a battery is 1.5%, because it is still within the ±2.0% dead zone, the system forces its deviation to zero and it does not participate in the current energy balance adjustment.

[0090] S107. Determine the basic balance duty cycle of each individual cell based on the balance demand deviation.

[0091] After identifying the balancing demand deviation of each individual battery cell, the control system needs to further quantify this deviation into specific control commands to drive the balancing circuit to perform current regulation. Therefore, in step S107, the corresponding basic balancing duty cycle for each individual battery cell is calculated based on its balancing demand deviation. The basic balancing duty cycle refers to the duty cycle (i.e., on-time ratio) at which the balancing circuit should operate under ideal conditions, without considering the effects of polarization risk and aging differences, to meet the current charge migration requirements. The core of this step lies in converting the energy imbalance representation of the deviation into input parameters for PWM control, thereby achieving an effective transition from decision logic to execution control. Specifically, it can include the following steps: If the balance demand deviation is zero, the basic balance duty cycle is directly set to zero to keep the balance circuit in the off state; if the balance demand deviation is not zero, the absolute value of the balance demand deviation is determined as the net balance demand amplitude; the smaller of the ratio of the net balance demand amplitude to the preset full-load reference value and the value in value one is determined as the normalized demand coefficient, where the preset full-load reference value is used to characterize the minimum deviation required for the balance circuit to enter the full-load working state; based on the normalized demand coefficient, affine mapping calculation is performed within the interval formed by the preset effective duty cycle threshold and the preset rated full-load duty cycle to obtain the basic balance duty cycle, where the effective duty cycle threshold is the minimum conduction ratio required for the balance circuit to maintain effective current output.

[0092] In implementing step S107, to achieve precise control over the equalization adjustment of each individual cell in the battery pack, the equalization demand deviation obtained in the previous stage needs to be converted into specific PWM control parameters, namely the basic equalization duty cycle. The core of this step is to construct a reasonable mapping relationship so that the SOC deviation can effectively drive the conduction time of the equalization circuit, thereby adjusting the intensity of the energy migration process.

[0093] For individual cells whose equalization demand deviation is forcibly set to zero, it means that their current state of charge (SOC) is already within the dynamic insensitivity dead zone, which falls within the system's allowable tolerance range for charge error. These cells do not need to participate in this round of energy migration; therefore, their basic equalization duty cycle is directly set to zero, ensuring that their corresponding equalization circuit is in the off state. This is done to avoid overreacting to small SOC deviations, prevent ineffective energy regulation and waste of control resources, and also reduce the risk of heat buildup.

[0094] For individual cells with a non-zero equilibrium demand deviation, the system first obtains the absolute value of this deviation, denoted as the net equilibrium demand amplitude. This amplitude represents the actual degree of charge deviation between the cell and the group equilibrium benchmark, and serves as the starting point for subsequent mapping calculations. To standardize this deviation value, the system presets a full-load benchmark value, which represents the minimum SOC deviation threshold required by the equilibrium circuit under full load. If the net deviation amplitude exceeds this benchmark value, the equilibrium circuit will adjust at maximum capacity; if it is below this value, proportional scaling is required.

[0095] The system calculates the ratio of the net equilibrium demand amplitude to the preset full-load reference value, compares this ratio with the value 1, and takes the smaller value as the normalized demand coefficient. This coefficient ranges from 0 to 1, reflecting the relative position of the current required adjustment intensity within the full-load range, exhibiting good normalization and control linearity. This normalized demand coefficient is used as input to further calculate the duty cycle of the PWM control.

[0096] Considering the actual operating characteristics of the balancing circuit, the system does not allow unlimited adjustment of the PWM duty cycle starting from 0%. Instead, an effective duty cycle threshold is set, representing the minimum conduction ratio required for the balancing circuit to maintain effective current output. The system also sets a rated full-load duty cycle, representing the upper conduction limit during maximum deviation adjustment. Based on these two boundary values, the control system constructs an affine mapping function that linearly maps the normalized demand coefficient from the interval [0,1] to the interval [effective threshold, upper duty cycle limit], thereby obtaining the basic balancing duty cycle of the target battery. This mapping ensures that the system has adaptive adjustment capability for different levels of energy deviation, while avoiding ineffective conduction at low duty cycles and overload risks at high duty cycles.

[0097] Specifically, the affine mapping calculation is essentially a linear interval scaling operation, designed to map the normalized demand coefficients to the effective operating range of the equalization circuit. Its calculation formula is as follows: ,in, The aforementioned basic equilibrium duty cycle; The preset effective duty cycle threshold is (e.g., 10%). The preset rated full-load duty cycle (e.g., 90%). This is the normalized demand coefficient. Through this formula, the system ensures that even with the normalized demand coefficient... In extremely rare cases, the output duty cycle will not fall below the hardware activation threshold. This avoids the problem of MOSFETs operating in the nonlinear region or failing to conduct effectively.

[0098] For example, if a battery's equalization demand deviation is 1.5%, its full-load baseline is 2.0%, its effective threshold is set to 10%, and its rated full-load duty cycle is set to 90%, then the normalized demand coefficient is min(1.5 / 2.0,1) = 0.75. Through affine mapping, the basic equalization duty cycle can be obtained as: 10% + 0.75 × (90% − 10%) = 70%. This means that the battery will conduct its equalization circuit with a 70% PWM duty cycle to achieve an energy regulation intensity that matches its deviation level.

[0099] Through the above methods, the system not only achieves accurate conversion of SOC deviation into control signal, but also effectively ensures the response efficiency and safety boundary of the equalization process, laying a reliable foundation for subsequent duty cycle correction based on aging and polarization factors.

[0100] S108. Based on the polarization safety factor and the aging weighting factor, the basic equalization duty cycle is corrected to generate the target pulse width modulation duty cycle.

[0101] After completing the initial calculation of the basic equalization duty cycle, to further ensure the safety and adaptability of the equalization control process, the control system needs to dynamically correct the duty cycle based on the current health status and polarization risk of each individual cell. This is to avoid applying excessively high equalization current to cells with severe aging or abnormal polarization impedance, which could lead to thermal runaway or accelerated performance degradation. Therefore, in step S108, the system performs a weighted correction on the basic equalization duty cycle of each individual cell based on the polarization safety factor and aging weighting factor calculated in the previous steps, thereby generating the target pulse width modulation duty cycle for actual PWM control. This correction process dynamically adjusts the upper limit of the duty cycle output capability through the superposition of weighting factors, making the current distribution more suitable for individual battery differences and significantly improving the safety boundary of the entire battery pack operation while ensuring equalization efficiency. Specifically, it can include the following steps: multiplying the polarization safety coefficient and the aging weighting coefficient to obtain a comprehensive state correction factor, which is used to characterize the current carrying capacity weight of the individual battery under the constraints of current polarization thermal accumulation risk and aging degree; calculating the product of the basic equalization duty cycle and the comprehensive state correction factor to obtain the theoretical corrected duty cycle; if the theoretical corrected duty cycle is less than the preset effective duty cycle threshold, forcing the target pulse width modulation duty cycle to zero; if the theoretical corrected duty cycle is greater than or equal to the preset effective duty cycle threshold, determining the smaller value between the theoretical corrected duty cycle and the preset hardware safety limit duty cycle as the target pulse width modulation duty cycle.

[0102] In practical implementation, to reliably correct the basic equilibrium duty cycle of individual cells, it is necessary to comprehensively consider the limitations of the current carrying capacity of the cell's current polarization safety state and aging degree. Therefore, the system multiplies the polarization safety factor and aging weighting factor obtained for each individual cell in the previous steps to obtain a new parameter called the comprehensive state correction factor. This factor is used to quantify the maximum equilibrium adjustment intensity that the current individual cell can accept under multiple state constraints. Here, the polarization safety factor reflects the risk of thermal accumulation caused by polarization impedance fluctuations under current operating conditions; the smaller the value, the less safe it is. The aging weighting factor, on the other hand, is an aging intensity index based on the ohmic internal resistance growth rate and temperature conditions, used to suppress excessive energy migration in aging cells. The result of multiplying the two factors forms a comprehensive weighted assessment of the target cell's "health carrying capacity".

[0103] The system multiplies the base equilibrium duty cycle calculated in the previous stage with the comprehensive state correction factor to obtain a new control quantity, called the theoretical corrected duty cycle. Essentially, this operation adaptively scales the original equilibrium control intent under state constraints to avoid excessive equilibrium adjustments on high-risk batteries, thereby improving the system's operational safety and long-term stability.

[0104] To ensure that the final output PWM signal has sufficient regulation effect without causing the equalization circuit to fail to conduct due to an excessively low duty cycle, the system sets a preset effective duty cycle threshold. This threshold is the minimum effective operating threshold value of the equalization circuit, determined experimentally. If the theoretically corrected duty cycle is lower than this threshold, the system forces it to zero, completely shutting down the equalization channel of the battery to avoid wasting resources and circuit losses due to ineffective conduction. When the theoretically corrected duty cycle is greater than or equal to this threshold, the system further compares it with a preset hardware safety limit duty cycle. This limit value is determined by hardware design parameters and represents the maximum allowable duty cycle of the equalization circuit within a safe range. Finally, the system takes the smaller of the theoretically corrected duty cycle and this limit value as the target pulse width modulation duty cycle for the individual battery and uses it to drive the corresponding equalization circuit for conduction control.

[0105] For example, if a single battery cell has a base equalization duty cycle of 80%, a polarization safety factor of 0.75, and an aging weighting factor of 0.6, then the comprehensive state correction factor is 0.75 × 0.6 = 0.45, resulting in a theoretically corrected duty cycle of 80% × 0.45 = 36%. If the effective duty cycle threshold is 10% and the hardware safety limit is 90%, then the final target PWM duty cycle is 36%. This process ensures that even if the original control intention is strong, it will be appropriately weakened when the battery state is abnormal, thereby achieving a safe and differentiated equalization adjustment strategy.

[0106] S109. The equalization circuit connected to both ends of each individual battery cell is turned on according to the target pulse width modulation duty cycle.

[0107] In step S109, the target pulse width modulation duty cycle, which has been corrected in the previous step, is applied to the specific equalization current adjustment operation. This allows for dynamic adjustment of the energy difference between individual cells with different states of charge within the target battery pack, while ensuring safety. To achieve this, the control system needs to precisely bind the target duty cycle corresponding to each individual cell to its connected equalization circuit in real time. This is then used as a control parameter to drive the pulse width modulation (PWM) signal output, thereby controlling the conduction duration and frequency of the equalization circuit in an adjustable manner, and indirectly controlling the equalization current intensity flowing through that individual cell.

[0108] In practical implementation, the system employs a high-speed digital controller such as a microcontroller unit (MCU) or a digital signal processor (DSP). The target duty cycle is used as input, and a PWM signal is generated through a timer module. Each PWM signal corresponds to the equalization branch of a single battery cell. The duty cycle of the PWM signal is the proportion of the conduction period within a complete cycle. The system applies this duty cycle to the drive terminals of MOSFETs or other switches to achieve on / off modulation of the equalization branch. The equalization circuit can employ an inductive active equalization structure or a capacitor-transfer equalization structure. The inductive type is more suitable for high-frequency PWM control. Its working principle is to establish a current path through the inductor during conduction and release energy to the target battery or other load through the inductor during turn-off, thereby achieving energy transfer.

[0109] By controlling the high-to-low level ratio of the PWM signal through the target duty cycle, the average current per unit time is essentially altered, thus controlling the equalization energy transfer rate per unit time. A higher target duty cycle increases the conduction time, leading to a higher average equalization current and a faster energy transfer rate; conversely, a lower target duty cycle weakens this effect. Since the previous steps have already corrected for aging and polarization states in the duty cycle, this PWM control not only meets the need for adjusting battery state-of-charge differences but also effectively avoids applying excessive current to severely aged or highly polarized batteries, thereby preventing localized overheating or side reactions.

[0110] Taking a specific implementation as an example, assuming the target pulse width modulation duty cycle of the third cell is 45%, the system controls the PWM controller of the corresponding equalization branch to output a PWM waveform with a period of 100 microseconds, where the high-level duration is 45 microseconds and the low-level duration is 55 microseconds. This PWM signal drives the MOSFET switch in the inductive equalization circuit to be on for 45% of the time in each cycle. The current is modulated to an average value of approximately 45% of the full-load equalization current during that cycle, thereby achieving a precise current regulation effect.

[0111] Through the above methods, the equalization control system realizes closed-loop current distribution control based on the target duty cycle, which not only improves the overall equalization efficiency of the battery pack, but also significantly enhances the system's adaptability and operational stability in the face of battery heterogeneity, aging evolution and thermal safety risks.

[0112] Please see Figure 3 This is a schematic diagram of the battery pack equalization control based on multidimensional state mapping in an embodiment of this application.

[0113] It should be noted that, Figure 3 The structure of a battery pack equalization control system based on multidimensional state mapping shown is merely an example and should not impose any limitations on the functionality and scope of application of the embodiments of the present invention.

[0114] like Figure 3 As shown, a battery pack balancing control system based on multi-dimensional state mapping includes a central processing unit 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory 302 or a program loaded from a storage section 308 into a random access memory 303, such as executing the methods described in the above embodiments. The random access memory 303 also stores various programs and data required for system operation. The central processing unit 301, the read-only memory 302, and the random access memory 303 are interconnected via a bus 304. An input / output interface 305 is also connected to the bus 304.

[0115] The following components are connected to the input / output interface 305: an input section 306 including audio input devices, push-button switches, etc.; an output section 307 including an LCD display, audio output devices, indicator lights, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the input / output interface 305 as needed. A removable medium 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 310 as needed so that computer programs read from it can be installed into the storage section 308 as needed.

[0116] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by central processing unit 301, it performs the various functions defined in the present invention.

[0117] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, flash memory, optical fiber, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0118] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0119] Specifically, a battery pack equalization control system based on multidimensional state mapping in this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, it implements a battery pack equalization control method based on multidimensional state mapping provided in the above embodiment.

[0120] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the battery pack balancing control system based on multidimensional state mapping described in the above embodiments; or it may exist independently and not incorporated into the battery pack balancing control system based on multidimensional state mapping. The storage medium carries one or more computer programs, which, when executed by a processor of the battery pack balancing control system based on multidimensional state mapping, cause the battery pack balancing control system based on multidimensional state mapping to implement the battery pack balancing control method based on multidimensional state mapping provided in the above embodiments.

Claims

1. A battery pack equalization control method based on multidimensional state mapping, characterized in that, The method includes: Acquire the terminal voltage data, operating current data, and surface temperature data of each individual cell in the target battery pack; Construct an equivalent circuit model for each of the individual cells, and calculate the ohmic internal resistance and polarization impedance of each individual cell based on the terminal voltage data, the operating current data, and the equivalent circuit model. Calculate the impedance growth rate of the ohmic internal resistance value relative to the preset nominal internal resistance value; Based on the impedance growth rate and the surface temperature data, the health status value of each individual cell is determined, and an aging weighting coefficient is determined based on the health status value of each individual cell. The polarization safety factor of each individual cell is determined based on the polarization impedance value; The state of charge (SOC) value of each individual cell in the target battery pack is determined based on a preset charge calculation model, and the balance demand deviation is determined based on the SOC value. The basic balance duty cycle of each individual cell is determined based on the balance demand deviation. The basic equalization duty cycle is corrected based on the polarization safety factor and the aging weighting factor to generate the target pulse width modulation duty cycle; The equalization circuit connected to both ends of each individual cell is turned on according to the target pulse width modulation duty cycle.

2. The method according to claim 1, characterized in that, The calculation of the ohmic internal resistance and polarization impedance of each individual cell based on the terminal voltage data, the operating current data, and the equivalent circuit model specifically includes: Monitor the time change rate of the operating current data, and determine the moment when the time change rate first exceeds the preset sudden change threshold as the excitation trigger moment; The difference between the first terminal voltage data at the first sampling time and the second terminal voltage data at the second sampling time in the terminal voltage data is calculated to obtain the transient voltage response value. Both the first sampling time and the second sampling time are adjacent to the excitation triggering time, and the first sampling time is before the excitation triggering time. Calculate the transient current surge amplitude between the first sampling time and the second sampling time based on the operating current data; The ratio of the transient voltage response value to the transient current change amplitude is determined as the ohmic internal resistance value; Obtain the continuous terminal voltage sampling sequence within a preset relaxation time window after the excitation triggering time; A polarization voltage response sequence is generated based on the continuous terminal voltage sampling sequence and the equivalent circuit model, and the polarization impedance value is calculated based on the polarization voltage response sequence.

3. The method according to claim 2, characterized in that, The process of generating a polarization voltage response sequence based on the continuous terminal voltage sampling sequence and the equivalent circuit model, and calculating the polarization impedance value based on the polarization voltage response sequence, specifically includes: Based on the equivalent circuit model, the open-circuit voltage state value corresponding to the excitation triggering time is determined. The difference between the target terminal voltage value and the non-polarized composite voltage value at each sampling point in the continuous terminal voltage sampling sequence is calculated respectively. The polarized voltage response sequence after removing the static voltage and ohmic voltage drop is obtained. The non-polarized composite voltage value is the sum of the product of the real-time current value corresponding to the sampling point and the ohmic internal resistance value and the open-circuit voltage state value. The voltage change amplitude of the polarization voltage response sequence within the relaxation time window is obtained, and the ratio of the voltage change amplitude to the transient current change amplitude is calculated to determine the polarization impedance value of the single cell.

4. The method according to claim 1, characterized in that, The determination of the polarization safety factor for each individual cell based on the polarization impedance value specifically includes: Based on the polarization impedance values ​​of all the individual cells, the average group polarization impedance of the target battery pack is calculated; Calculate the impedance dispersion ratio of each individual cell relative to the average polarization impedance of the group; When the impedance dispersion deviation ratio is greater than the preset thermal accumulation warning threshold, the single cell is determined to be in a polarization thermal accumulation risk state, and the ratio of the preset thermal accumulation warning threshold to the impedance dispersion deviation ratio is calculated to obtain the polarization safety factor. When the impedance discrepancy ratio is less than or equal to the preset thermal accumulation warning threshold, the single cell is determined to be in a polarization safety state, and the preset reference coefficient is determined as the polarization safety factor.

5. The method according to claim 1, characterized in that, The step of determining the equilibrium demand deviation based on the state of charge value specifically includes: The arithmetic mean of the target battery pack is calculated based on the state of charge values ​​of all the individual cells to obtain the group equilibrium benchmark value, and the original difference between the state of charge value of each individual cell and the group equilibrium benchmark value is calculated. The average of the absolute values ​​of the original differences of all the individual cells is calculated to obtain the group discrete characteristic value, which is used to characterize the statistical average discrete amplitude of the state of charge distribution of each individual cell in the target battery pack. The dead zone radius is determined based on the discrete characteristic values ​​of the population, and the dead zone radius is positively correlated with the discrete characteristic values ​​of the population. A dynamic insensitive dead zone is constructed based on the dead zone radius. The dynamic insensitive dead zone includes a positive cutoff boundary value and a negative cutoff boundary value. The absolute values ​​of the positive cutoff boundary value and the negative cutoff boundary value are both the dead zone radius. If the original difference is within the dynamic insensitivity dead zone, it is determined that the individual cell is within the preset equalization allowable error tolerance range, and the equalization requirement deviation of the individual cell is forcibly set to zero; If the original difference is greater than the positive cutoff boundary value, the difference between the original difference and the positive cutoff boundary is determined as the equilibrium demand deviation degree. If the original difference is less than the negative cutoff boundary value, the difference between the original difference and the negative cutoff boundary value is determined as the equilibrium demand deviation.

6. The method according to claim 1, characterized in that, The determination of the basic balance duty cycle for each individual cell based on the balance demand deviation specifically includes: If the balance demand deviation is zero, the basic balance duty cycle is set directly to zero to keep the balance circuit in the off state. If the equilibrium demand deviation is not zero, then the absolute value of the equilibrium demand deviation is determined as the net equilibrium demand magnitude. The smaller of the ratio of the net balance demand amplitude to the preset full load reference value and the value in Value 1 is determined as the normalized demand coefficient. The preset full load reference value is used to characterize the minimum deviation required for the balance circuit to enter the full load working state. Based on the normalized demand coefficient, affine mapping calculation is performed within the interval formed by the preset effective duty cycle threshold and the preset rated full-load duty cycle to obtain the basic equalization duty cycle. The effective duty cycle threshold is the minimum conduction ratio required for the equalization circuit to maintain effective current output.

7. The method according to claim 6, characterized in that, The step of correcting the basic equalization duty cycle based on the polarization safety factor and the aging weighting factor to generate the target pulse width modulation duty cycle specifically includes: Multiplying the polarization safety factor by the aging weighting factor yields the comprehensive state correction factor, which is used to characterize the current carrying capacity weight of the single cell under the constraints of current polarization thermal accumulation risk and aging degree. The theoretical corrected duty cycle is obtained by multiplying the basic equilibrium duty cycle by the comprehensive state correction factor. If the theoretically corrected duty cycle is less than the preset effective duty cycle threshold, the target pulse width modulation duty cycle will be forcibly set to zero; If the theoretically corrected duty cycle is greater than or equal to the preset effective duty cycle threshold, the smaller value between the theoretically corrected duty cycle and the preset hardware safety limit duty cycle is determined as the target pulse width modulation duty cycle.

8. A battery pack balancing control system based on multi-dimensional state mapping, characterized in that, The battery pack balancing control system based on multidimensional state mapping includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the battery pack balancing control system based on multidimensional state mapping to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the battery pack balancing control system based on multidimensional state mapping, the battery pack balancing control system based on multidimensional state mapping performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the battery pack balancing control system based on multidimensional state mapping, the battery pack balancing control system based on multidimensional state mapping performs the method as described in any one of claims 1-7.

Images