Active equalization control method and device for lithium ion battery pack

Abstract

This invention discloses an active balancing control method and device for lithium-ion battery packs, relating to the field of battery charging and discharging. The method first establishes a balancing priority based on the health state, sets an initial range threshold and an upper limit for the number of refreshes, and then uses the balancing priority to adjust the decision state of charge and construct an energy transfer path matrix to perform active balancing. Based on real-time monitoring of the state of charge range, when the number of balancing executions reaches the upper limit, an aging factor calculated based on the maximum decrease in health state and the average decay rate of the state of charge range is introduced. Combined with a convergence factor calculated based on the current number of balancing executions, the upper limit for the number of refreshes and the range threshold are adaptively updated and reset. This invention, through quantification of balancing priorities and a dual dynamic adjustment mechanism based on the aging factor and the convergence factor, achieves accurate response to the aging characteristics of the battery pack, significantly improving the long-term stability and convergence efficiency of the balancing system.

Description

Technical Field

[0001] This invention relates to the field of battery charging and discharging, and in particular to an active balancing control method and device for lithium-ion battery packs. Background Technology

[0002] In lithium-ion battery pack applications, differences in manufacturing processes, operating environments, and aging levels among individual cells lead to inconsistent states of charge (SOC), affecting the overall performance and safety of the battery pack. Existing passive balancing technologies have low energy utilization rates, while active balancing technologies, although more efficient, still suffer from slow balancing speed, complex control, and the generation of intermediate micro-cycles during the balancing process, resulting in energy loss and temperature rise. Chinese patent CN115765086B discloses a time-optimal active balancing method for lithium-ion battery packs. This method establishes a relationship between battery SOC and equalizer PWM duty cycle by constructing a hierarchical active balancing topology, and employs a time-optimal model predictive control strategy to reduce micro-cycle cycles, thereby accelerating the balancing speed, reducing losses and temperature rise, and simplifying the system structure. However, after long-term operation of the battery pack, the aging degree and performance degradation of each individual cell vary. If SOC is used as the sole balancing target, cells in poor health may be subjected to excessive balancing current or frequent charge-discharge cycles, accelerating their aging process and affecting the overall lifespan and reliability of the battery pack. Furthermore, this method aims to force all individual cells to converge to the same point, lacking dead-zone control based on dynamic thresholds, which can easily lead to frequent system operations under small pressure differences. Therefore, the existing technology needs further improvement. Summary of the Invention

[0003] To address the aforementioned issues, this invention proposes an active equalization control method and device for lithium-ion battery packs. This method determines equalization priorities based on the health status to achieve differentiated equalization protection for aging batteries, and constructs a closed-loop adaptive control system based on range feedback to dynamically update parameters, thereby significantly extending battery pack lifespan, avoiding over-equalization oscillations, and adapting to changes in operating conditions throughout the battery's entire life cycle.

[0004] The objective of this invention can be achieved through the following technical solutions: An active balancing control method for lithium-ion battery packs includes the following steps: Step 1: Obtain the state of charge and health status of each individual cell in the battery pack, determine the balancing priority of each individual cell based on the health status, and set the initial range threshold and refresh count limit. Step 2: Determine the correction factor based on the charge-discharge characteristic curve of the battery pack, adjust the state of charge according to the balancing priority to obtain the decision state of charge, and determine the balancing center point based on the decision state of charge. Step 3: Construct the energy transfer path matrix between individual cells by combining the correction factor and the equilibrium center point, and control the energy transfer device to perform energy transfer; Step 4: Calculate the range of the current state of charge of each individual cell. If the range is less than the range threshold, end the task; otherwise, proceed to step 8. Step 5: If the number of balance executions reaches the refresh count limit, update the refresh count limit and the range threshold, reset the balance execution count to zero, and return to Step 1; otherwise, increment the balance execution count by 1 and proceed to Step 2. The upper limit for the number of updates and the range threshold include: reading the maximum value of the decline in the health status of a single battery cell and the average decay rate of the state of charge range; calculating the aging factor based on the maximum value and the average decay rate; calculating the convergence factor based on the current number of equalization executions; and updating the upper limit for the number of updates and the range threshold based on the aging factor and the convergence factor.

[0005] In this invention, in step 1, determining the equalization priority of each individual battery cell based on its health status includes: determining the first priority of each individual battery cell as an energy outputter and the second priority as an energy receiver.

[0006] In this invention, in step 2, the correction factor is determined based on the position of the battery pack's state of charge on the charge-discharge characteristic curve: during charging, when the state of charge is in the initial polarization region or the final polarization region, the correction factor takes a fixed negative value; when the state of charge is in the stable region, the correction factor changes linearly with the state of charge. During discharging, when the state of charge is in the initial polarization region or the final polarization region, the correction factor takes a fixed positive value; when the state of charge is in the stable region, the correction factor changes linearly with the state of charge.

[0007] In this invention, in step 2, the decision states of charge are sorted by numerical value, and the decision states of charge corresponding to the median are used as the initial center point. The sum of the absolute deviations between the decision states of charge of each individual cell and the initial center point is calculated. The initial center point is shifted in the direction with higher decision states of charge density. After each shift, the sum of absolute deviations is recalculated until the sum of absolute deviations no longer decreases. The center point at this time is determined as the equilibrium center point.

[0008] In this invention, in step 3, the upper limit of the equilibrium threshold Q1 = Q0 + [(1 + f)δ] is calculated, and the lower limit of the equilibrium threshold Q2 = Q0 - [(1 + f)δ] is calculated, where f is a correction factor, Q0 is the equilibrium center point, and δ is the basic equilibrium threshold. Individual cells with a decision state of charge higher than the upper limit of the equilibrium threshold are marked as discharge groups, and individual cells with a state of charge lower than the lower limit of the equilibrium threshold are marked as charging groups.

[0009] In this invention, in step 3, an energy transfer demand matrix is ​​constructed with the output energy of each individual battery in the discharge group as the row constraint and the demand energy of each individual battery in the charging group as the column constraint. The output energy is determined based on the difference between the current state of charge of the individual battery in the discharge group and the lower limit of the equilibrium threshold, and the demand energy is determined based on the difference between the upper limit of the equilibrium threshold and the current state of charge of the individual battery in the charging group.

[0010] In this invention, in step 3, with the goal of maximizing the sparsity of the energy transfer demand matrix, the single battery with the largest output energy in the discharge group is preferentially matched to the single battery with the largest energy demand in the charging group. This process is iterated until all energy demands are met or all output energy is exhausted. An energy transfer path matrix is ​​generated based on the matching results. The non-zero elements in this energy transfer path matrix represent the amount of energy transferred between the single battery in the discharge group and the single battery in the charging group.

[0011] In this invention, in step 3, the energy transfer device is any one of an isolated DC-DC converter, a non-isolated DC-DC converter, or an inductive equalizer.

[0012] In this invention, in step 5, a basic range threshold is determined based on the aging factor between a preset initial range threshold and a maximum allowable range threshold. The basic range threshold is then adjusted based on the convergence factor to obtain an updated range threshold. The updated range threshold is not lower than a preset protection threshold.

[0013] A control device for implementing the active balancing control method of the lithium-ion battery pack, comprising: The data acquisition module is used to acquire the state of charge and health status of each individual battery cell in real time. The data processing module is used to determine the balancing priority of each individual cell based on its health status, adjust the state of charge based on the balancing priority, obtain the decision state of charge, and determine the balancing center point. The data analysis module is used to determine the correction factor based on the charge and discharge characteristic curves, determine the upper and lower limits of the balance threshold in combination with the balance center point, and mark the individual cells with a decision state of charge higher than the upper limit of the balance threshold as the discharge group and those with a state of charge lower than the lower limit of the balance threshold as the charging group. The path planning module is used to construct the energy transfer demand matrix between the discharge group and the charging group, and generate the energy transfer path matrix. An energy transfer device for transferring energy from individual cells in a discharge group to individual cells in a charging group according to the energy transfer path matrix; The status monitoring module is used to calculate the current range of the state of charge of each individual battery cell, determine whether to end the equalization process, and determine whether the number of equalization executions has reached the refresh limit, and decide whether to recalculate the equalization priority and update the range threshold and refresh limit. The parameter update module is used to update the range threshold and the upper limit of refresh count.

[0014] The present invention provides an active balancing control method and device for lithium-ion battery packs, which has the following beneficial effects: First, by introducing a health state to determine the balancing priority and adjusting the decision state of charge, the present invention achieves a differentiated balancing strategy where healthy cells discharge more and aging cells discharge less, fundamentally avoiding accelerated degradation of aging cells due to frequent charging and discharging, and significantly extending the overall lifespan of the battery pack. Second, a closed-loop adaptive control system based on range feedback is constructed to determine the balancing termination time in real time, avoiding over-balancing and system oscillation, and improving the stability of the balancing process. Furthermore, the method determines whether to update the refresh count limit and range threshold based on the refresh count limit, so that the balancing strategy matches the current aging level of the battery pack. By introducing an aging factor and a convergence factor to dynamically update the range threshold and refresh count limit, and setting a lower limit protection mechanism, the balancing strategy can adapt to changes in operating conditions throughout the battery's entire life cycle, avoiding ineffective balancing and wasted computational resources. Attached Figure Description

[0015] Figure 1 A schematic diagram illustrating effective active balancing of a lithium-ion battery pack; Figure 2 A schematic diagram illustrating the insufficient active balancing of a lithium-ion battery pack; Figure 3 This is a flowchart of the active balancing control method for lithium-ion battery packs of the present invention; Figure 4 This is a control circuit diagram of the energy transfer device of the present invention; Figure 5 This is a diagram showing the voltage changes of each individual cell during a single active balancing process according to the present invention. Figure 6 This is a charge-discharge characteristic curve of the lithium-ion battery pack of the present invention; Figure 7 This is a schematic diagram of the control device of the present invention. Detailed Implementation

[0016] To better implement the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0017] like Figure 1As shown, active balancing control can integrate the energy states within and between lithium-ion battery packs. In existing technologies, lithium-ion battery pack balancing strategies often employ fixed thresholds. Overly strict thresholds during the voltage plateau period lead to frequent ineffective balancing, while overly broad thresholds at the end of charge / discharge result in insufficient balancing. Figure 2 As shown. Furthermore, the direct "highest to lowest" matching method involves numerous switching actions, resulting in significant energy loss, and it does not consider differences in battery health status, thus accelerating the degradation of aging batteries.

[0018] This invention prioritizes balancing based on the battery's health status, minimizing discharge and maximizing charge in aging batteries. It dynamically determines correction factors based on charge-discharge characteristic curves, establishes a balancing center point based on the adjusted state of charge, and defines upper and lower balancing thresholds, marking discharge and charging groups. By constructing an energy transfer demand matrix between discharge and charging groups, and planning an energy transfer path matrix with maximum matrix sparsity as the objective, the invention controls the energy transfer device to perform active balancing. Furthermore, the invention calculates the state of charge range and compares it with a range threshold. If the range is not met and the refresh count limit is reached, an aging factor calculated based on the maximum health status decline and the average decay rate of the range, along with a convergence factor calculated based on the number of balancing executions, is introduced. The range threshold and refresh count limit are dynamically updated, and active balancing is re-executed. This invention protects aging batteries through a health status-based priority strategy, avoids over-balancing and oscillations through closed-loop adaptive control based on range feedback, and dynamically adapts to all lifecycle operating conditions, significantly extending the overall battery pack lifespan and improving balancing stability and robustness. Example 1

[0019] Reference Figures 3 to 5 The active balancing control method for lithium-ion battery packs described in detail in this embodiment includes the following steps: Step 1: Obtain the state of charge (SOC) and health status of each individual cell in the battery pack. Determine the balancing priority of each individual cell based on its health status, and set an initial range threshold and refresh count limit. Determining the balancing priority of each individual cell based on its health status includes: determining a first priority for each individual cell as an energy output and a second priority as an energy receiver. Cells with lower health status have lower first priority and higher second priority. If priorities are the same, they are sorted by cell number. The typical range for the range threshold is 0.5% to 5%, and the typical range for the refresh count limit is 3 to 12. Specific values ​​are set according to battery pack consistency requirements, balancing hardware response speed, or system real-time requirements. In this embodiment, the lithium-ion battery pack used is a brand new battery pack with good initial consistency; therefore, the initial range threshold is set to, for example, 1%, and the refresh count limit is set to, for example, 4.

[0020] The health status is obtained through one of the following methods: measuring the internal resistance of a single battery cell and determining it based on the mapping relationship between internal resistance and health status; measuring the ratio of the current maximum usable capacity to the nominal capacity of a single battery cell; jointly estimating based on an equivalent circuit model combined with a Kalman filter algorithm; or determining it based on the fitting relationship between the time required for a specific voltage range during charging and discharging and the health status. In this embodiment, the health status is jointly estimated online based on an equivalent circuit model combined with a Kalman filter algorithm. The health status value ranges from 0-100%, with lower values ​​indicating more severe aging. Therefore, the single battery cell with the lowest health status is assigned the lowest discharge priority (first priority) and the highest charging priority (second priority) to protect the aging battery. In this embodiment, the nominal voltage of a single battery cell is 3.6V, and the rated capacity is 20Ah.

[0021] Step 2: Determine the correction factor based on the battery pack's charge-discharge characteristic curve, adjust the state of charge (SOC) according to the balancing priority to obtain the decision SOC, and determine the balancing center point based on the decision SOC. (Refer to...) Figure 6 In this embodiment, the charge-discharge characteristic curve uses the state of charge (SOC) of the battery pack as the abscissa and the voltage as the ordinate. The correction factor is determined based on the position of the battery pack's SOC on the charge-discharge characteristic curve: during charging, when the battery pack's SOC is in the initial polarization region or the final polarization region, the correction factor f takes a fixed negative value, for example, f = -0.1; when the battery pack's SOC is in the stable region, the correction factor changes linearly with the battery pack's SOC. During discharging, when the battery pack's SOC is in the initial polarization region or the final polarization region, the correction factor f takes a fixed positive value, for example, f = 0.1; when the battery pack's SOC is in the stable region, the correction factor changes linearly with the battery pack's SOC. The decision states of charge (DCCs) are sorted by numerical value. The DCC corresponding to the median is used as the initial center point. The sum of the absolute deviations between the DCC of each individual cell and the initial center point is calculated. The initial center point is then shifted in the direction with higher DCC density. After each shift, the sum of the absolute deviations is recalculated until the sum of the absolute deviations no longer decreases. The center point at this point is determined as the equilibrium center point. See Example 2 for details.

[0022] Step 3: Determine the upper and lower limits of the equalization threshold by combining the correction factor and the equalization center point. The upper limit of the equalization threshold Q1 = Q0 + [(1 + f)δ], and the lower limit of the equalization threshold Q2 = Q0 - [(1 + f)δ], where f is the correction factor, Q0 is the equalization center point, and δ is the basic equalization threshold used to control the basic width of the equalization dead zone. The basic equalization threshold δ is determined based on the initial consistency of the battery pack or the current SOC range. In this embodiment, δ is taken as 1 / 4 of the current SOC range of the battery pack, and the value range of δ is limited to 1%~5%. If the calculated value is less than 1%, then 1% is used; if it is greater than 5%, then 5% is used.

[0023] Step 4: Mark individual cells with a decision state of charge (SBC) above the upper limit of the balancing threshold as belonging to the discharge group, and those below the lower limit as belonging to the charge group. Since severely aged cells are adjusted downwards during discharge (lowering their first priority) and upwards during charging (increasing their second priority), aged cells are more likely to be assigned to the charge group and less likely to be assigned to the discharge group, thus protecting aged cells. If their SBC is between the upper and lower limits of the balancing threshold, they are in the balancing dead zone and do not participate in this round of active balancing.

[0024] Step 5: Construct an energy transfer demand matrix between the discharge group and the charging group. With the objective of maximizing matrix sparsity, determine the energy transfer amount between individual cells, resulting in an energy transfer path matrix. Using the output energy of each individual cell in the discharge group as row constraints and the demand energy of each individual cell in the charging group as column constraints, construct an energy transfer demand matrix with elements as non-negative decision variables to be optimized. The output energy is determined based on the difference between the current state of charge (SOC) of the individual cells in the discharge group and the lower limit of the equilibrium threshold, while the demand energy is determined based on the difference between the upper limit of the equilibrium threshold and the current SOC of the individual cells in the charging group. With the objective of maximizing matrix sparsity, prioritize matching the individual cell with the largest output energy in the discharge group to the individual cell with the largest demand energy in the charging group. Iterate this process until all demand energy is satisfied or all output energy is exhausted. Generate an energy transfer path matrix based on the matching results. The non-zero elements in this energy transfer path matrix represent the energy transfer amount between individual cells in the discharge group and individual cells in the charging group.

[0025] Step 6: Based on the energy transfer path matrix, control the energy transfer device to transfer energy from the individual cells in the discharge group to the individual cells in the charging group. The energy transfer device can be any one of an isolated DC-DC converter, a non-isolated DC-DC converter, or an inductive equalizer. The non-isolated DC-DC converter has a relatively simple structure and high energy transfer efficiency; the inductive equalizer relies on the energy storage and release characteristics of the inductor to transfer energy between battery cells. An isolated DC-DC converter (achieving electrical isolation through a transformer) can be applied to scenarios with high safety requirements.

[0026] Reference Figure 4 The energy transfer device uses a capacitive equalizer. For example, a lithium-ion battery pack has N individual cells, labeled O1, O2, ..., O... N The energy transfer device includes capacitors C1, C2, ... C N Inductors L1, L2, ... L N and current nodes I1, I2, ... I N Each channel contains individual cells O1, O2, ... O N Connected to a set of capacitors, inductors, and current nodes, the electrical interconnection of all current nodes forms a common energy bus. When it is necessary to transfer energy from a single cell m in the discharge group to a single cell r in the charging group, the path planning module simultaneously controls the switching timing of the channels containing single cells m and r, allowing the balancing current to flow between the two single cells through the common energy bus, thereby completing the energy transfer. Within each transfer cycle, the duty cycle or switching frequency of the PWM signal is adjusted to match the accumulated transferred energy with the energy transfer amount determined in the energy transfer path matrix. All energy transfer operations are executed sequentially according to the non-zero elements in the energy transfer path matrix until all are completed. Figure 5 In this embodiment, N is, for example, 5, and the lithium-ion battery pack has 5 individual cells, namely O1, O2, O3, O4, and O5.

[0027] Step 7: Calculate the range of the current state of charge (SOC) of each individual battery cell. If the range is less than the range threshold, end the task; otherwise, proceed to Step 8. The range refers to the difference between the maximum and minimum SOC values ​​of all individual batteries in the current battery pack. During the operation of an electric vehicle or energy storage system, the battery pack is not statically waiting for active balancing. While active balancing is in progress, the system may continue charging and discharging, causing the SOC range to constantly change. Furthermore, within a single active balancing cycle, the energy transfer device is limited by the balancing current (usually much smaller than the main circuit charging and discharging current). If the battery pack is simultaneously charging or discharging, the main circuit current will push up or pull down the SOC of all batteries, and its rate of change far exceeds the correction capability of the balancing device. This causes the adjustment effect of active balancing to be masked by the "overall shift" of charging and discharging, preventing the range from converging below the range threshold within a single active balancing cycle. Therefore, it is necessary to continuously assess the SOC range. If the range is less than the range threshold, it indicates that the current consistency of the battery pack meets the requirements, and the current active balancing task is terminated. However, the battery management system will continue to monitor the state of charge of all individual cells, and the balancing execution count will not be reset to zero until the upper limit of the number of balancing executions and the range threshold are updated. In this embodiment, the initial range threshold is 1%, that is, when the difference in the state of charge of all individual cells does not exceed 1%, the balancing is considered to have met the standard, and the entire balancing process is terminated. If the range is greater than 1%, it indicates that the inconsistency within the battery pack is too large, and active balancing needs to continue, proceeding to step 8.

[0028] Step 8: If the number of balancing executions reaches the refresh count limit, update the refresh count limit and the range threshold, reset the balancing execution count to zero, and return to Step 1. Otherwise, increment the balancing execution count by 1 and proceed to Step 2. When the number of balancing executions reaches the refresh count limit, it indicates that the battery pack has undergone sufficient active balancing, and its health status may have changed. At this point, return to Step 1 to reacquire the health status of each individual battery cell, update the balancing priority, and ensure the adaptability of subsequent balancing strategies.

[0029] Furthermore, the range threshold determines the termination condition of the active balancing process. If the range threshold is set too small, the active balancing process will be difficult to converge, leading to a significant increase in system losses; if the range threshold is set too large, the consistency between individual cells within the battery pack will be difficult to guarantee, and the capacity utilization rate will decrease accordingly. Moreover, as the battery pack ages, the capacity decay and internal resistance increase of each individual cell show a differentiated trend, and the dispersion of the available capacity window will intensify. The range threshold should dynamically change with the aging of the battery pack. The upper limit of the refresh count determines the calculation frequency of the balancing priority. The health status changes slowly, and under normal circumstances, frequent reassessment is not required. However, when the health status of a certain individual cell shows accelerated decay, it indicates that the current balancing priority is no longer suitable for the battery status, and the health status needs to be reassessed and the balancing priority updated. The updating of the upper limit of the refresh count and the range threshold includes: reading the maximum value of the decline in the health status of the individual cell and the average decay rate of the state of charge range; calculating the aging factor based on the maximum value and the average decay rate; calculating the convergence factor based on the current number of balancing executions; and updating the upper limit of the refresh count and the range threshold based on the aging factor and the convergence factor. For a preferred method for updating the range threshold and the upper limit of the refresh count, please refer to Embodiment 4. Example 2

[0030] This embodiment further illustrates the method of determining the correction factor in step 2, adjusting the state of charge according to the equilibrium priority, and then determining the optimal equilibrium center point based on the decision state of charge.

[0031] Determine the correction factor. The correction factor is determined based on the position of the battery pack's state of charge (SOC) on the charge-discharge characteristic curve. (Refer to...) Figure 6 The charge-discharge characteristic curve is divided into three regions: the initial polarization region, the stable region, and the final polarization region. When the state of charge (SOC) of the battery pack is in the initial or final polarization region of the charge-discharge characteristic curve, the battery voltage changes drastically, and the inconsistency between individual cells has the greatest impact on battery performance. In this case, the correction factor is taken as a fixed negative or positive value depending on whether the battery pack is in the charging or discharging state, so as to narrow the upper or lower limit of the equalization threshold. That is, the correction factor f is, for example, -0.1 or 0.1. When the SOC of the battery pack is in the stable region of the charge-discharge characteristic curve, the battery voltage changes slowly, and the inconsistency between individual cells is not easily amplified. In this case, the correction factor changes linearly with the SOC of the battery pack, that is, the correction factor f = -(Q4 - Q3) / 2I, where Q3 is the initial charge of the stable region, I is the charge range of the stable region, and Q4 is the SOC of the battery pack.

[0032] The state of charge (SOC) is adjusted based on the balancing priority. For each individual cell, a comprehensive correction coefficient k = (P1 - P2) / (P1 + P2) is calculated based on its first priority as an energy outputter (P1) and its second priority as an energy receiver (P2). When k is less than 0, the individual cell is more suitable as an energy receiver, and its SOC is increased to make it easier to be selected for the charging group. When k is greater than 0, the individual cell is more suitable as an energy outputter, making it easier to be selected for the discharging group. The decision SOC is Q' = Q + λk, where Q is the SOC of the individual cell before adjustment, and λ is the adjustment magnitude coefficient. The preferred value of λ is 0.01 to 0.05. λ can be dynamically adjusted according to the health status distribution of the battery pack, taking a larger value when the difference in health status between individual cells is large, and a smaller value when the difference is small.

[0033] The equilibrium center point is determined based on the decision state of charge (DSC). First, the DSCs of each individual cell are arranged in ascending order. The DSC corresponding to the median is used as the initial center point. A one-dimensional search method is then employed to determine the equilibrium center point. Specifically, the mean of the DSCs of each individual cell is calculated, and this mean is used as the initial center point. The initial center point is then shifted towards a direction with higher DSC density, and the sum of the absolute deviations at this initial center point is calculated. Where μ is the energy transfer efficiency, Δ1 is the dynamic threshold radius, Δ1=(1+f)δ, Q i 'For decision-making, the state of charge is greater than The decision state of charge (Q) of any single cell. j 'For decision-making, the state of charge is less than The decision state of charge of any single cell is determined. The process continues until the sum of the absolute deviations no longer decreases, at which point the center point is determined as the equilibrium center point Q0. Example 3

[0034] This embodiment further discloses an optimal method for constructing the energy transfer demand matrix between the discharge group and the charging group in step 5, and for planning the energy transfer path matrix with the goal of maximizing matrix sparsity. The energy transfer demand matrix is ​​constructed based on the actual state of charge (SOC) of each individual battery cell. The output energy E is determined based on the difference between the SOC of each individual battery cell in the discharge group and the lower limit of the equalization threshold Q2. The required energy D is determined based on the difference between the upper limit of the equalization threshold Q1 and the SOC of each individual battery cell in the charging group. That is, E m =Q m - Q2, D r =Q1-Q r Where m = 1, 2, ..., M, r = 1, 2, ..., R, M is the number of individual cells in the discharge group, R is the number of individual cells in the charging group, and E m Q represents the output energy of a single cell m in the discharge group.m Let D be the state of charge of a single cell m. r Q represents the energy required by a single cell r in the discharge group. r Let r be the state of charge of a single cell. Sort the cells in the charging group by their required energy from smallest to largest. Similarly, sort the cells in the discharging group by their output energy from smallest to largest, constructing an output energy vector E = [E1, E2, ..., E]. m ,…, E M ] T And the energy demand vector D=[D1, D2,…D r ,…, D R ] T Where E1≤E2≤…≤E m ≤…≤E M D1≤D2≤…≤D r ≤…≤D R Reconstruct the energy transfer demand matrix. The element x in this energy transfer demand matrix m,r This represents the amount of electricity that cell m is prepared to transfer to cell r, which is a non-negative decision variable (unknown) to be optimized.

[0035] Plan the energy transfer path matrix. Using the energy transfer demand matrix as a constraint, plan the energy transfer path matrix with the objective of maximizing its sparsity, i.e., minimizing the number of switching actions during the energy transfer process. The constraint is that the sum of all elements in each row cannot exceed the output energy of the corresponding discharge battery in that row, and the sum of all elements in each column cannot exceed the energy requirement of the corresponding rechargeable battery in that column. This embodiment uses a greedy matching strategy to plan the energy transfer scheme: sequentially matching the single battery with the largest output energy to the single battery with the largest energy requirement, with an energy transfer amount t. mr =min(E m D r The remaining energy of both parties is updated. The above operation is repeated until all required energy is met or all output energy is exhausted. Based on the matching result, an energy transfer path matrix V2 is generated. The non-zero elements in the energy transfer path matrix V2 represent the actual energy value transferred from the individual cells of the discharge group to the individual cells of the charging group. The matrix has the fewest non-zero elements, that is, the matrix sparsity is the largest. Example 4

[0036] This embodiment further discloses a preferred method for updating the range threshold and the upper limit of refresh counts in step 8. This embodiment introduces an aging factor and a convergence factor. The aging factor is used to quantify the overall aging degree of the battery pack and the difficulty of equalization, while the convergence factor is used to characterize the convergence progress of the equalization process. Together, they guide the adaptive updating of the range threshold and the upper limit of refresh counts.

[0037] Calculate the aging factor. The battery pack in this embodiment consists of N individual cells, and the state of health (SOH) of the nth individual cell is... n The maximum value of the decline in the health status of each individual cell The state-of-charge range before the current active balancing is SOC1, and the state-of-charge range after the current active balancing is SOC2. Therefore, the single-round range attenuation rate λ1 = (SOC1 - SOC2) / SOC1. The average of the single-round range attenuation rates from the most recent multiple rounds (e.g., 5 rounds) of active balancing is calculated to obtain the average attenuation rate λ2 of the current state-of-charge range. Then, the aging factor α = w1SOH max +w2(1-λ2), where w1 and w2 are preset weight coefficients that satisfy w1+w2=1, SOH max The larger the value or the lower the average degradation rate, the more severe the battery pack aging and the greater the difficulty in active balancing, thus the larger the aging factor.

[0038] Calculate the convergence factor. The convergence factor characterizes the convergence progress of the active equilibrium process and decreases monotonically with the number of equilibrium executions. Where K is the current number of times the load balancer is executed, K max β is the preset global maximum allowed refresh count limit (e.g., 10). min The preset minimum convergence factor ranges from 0.3 to 0.7. When active balancing just begins, the convergence factor is close to 1; as the number of executions increases, the convergence factor decreases linearly until it reaches the lower limit β. min .

[0039] Update the range threshold. Determine the baseline range threshold SOC based on the aging factor. base =SOC0+α(SOC max -SOC0), where SOC0 is the initial range threshold (corresponding to the standard threshold for a brand new battery pack), with a value ranging from 0.5% to 2%. max The maximum permissible range threshold (corresponding to the tolerance threshold for severely aged battery packs) ranges from 5% to 10%. Therefore, the updated range threshold SOC3 = βSOC. base The range threshold widens as the aging factor increases and narrows as the convergence factor decreases, adapting to aging differences while ensuring convergence to a reasonable range in the later stages of equalization. To prevent the active equalization from failing to terminate due to an excessively strict range threshold when the battery is severely aged, the updated range threshold should also satisfy SOC3 = max(βSOC). base SOC limit ), where SOC limit This refers to the protection threshold for aging batteries.

[0040] Update refresh count limit. Determine the base refresh count limit K based on the aging factor. base =K min +α(K max -K min ), where K min The minimum allowed number of refreshes (e.g., 4) corresponds to a brand new battery pack or easy active balancing, and the value ranges from 3 to 5, K. max If the value ranges from 10 to 15, then the maximum number of refreshes after the update is... ,in, This indicates rounding down. The upper limit of refresh count increases with the aging factor, giving the system more opportunities for adjustment; however, it decreases with the convergence factor, avoiding waste of computing resources or indefinite continuation of the active balancing process due to excessive refresh counts.

[0041] Furthermore, in this embodiment, those skilled in the art can adjust the values ​​of w1 and w2 according to actual operating conditions. For example, when the balancing effect fluctuates significantly, w2 can be appropriately reduced to minimize the impact of a single fluctuation. Because in this embodiment, the overall aging degree of the battery pack typically has a greater impact on the balancing difficulty than the fluctuation of a single balancing effect, the preset weighting coefficients are w1 = 0.6 and w2 = 0.4. That is, the maximum health status decrease is SOH. max The weight of the variable is higher than the weight of the average degradation rate λ2. In another embodiment, the weighting coefficients w1 and w2 are dynamically determined based on the historical equalization data of the battery pack, so that the index with greater fluctuation is assigned a smaller weighting coefficient, thereby reducing the interference of outliers on the aging factor. Specifically, the SOH during the past T equalization processes (e.g., T=10) is recorded. max Calculate SOH using the values ​​of λ and λ2 respectively. max Given the variances σ1² and σ2² of λ², we have w1 = σ2² / (σ1² + σ2²) and w2 = σ1² / (σ1² + σ2²). When σ1² = σ2² = 0, w1 and w2 take the value of 0.5. Example 5

[0042] Reference Figure 7This embodiment discloses a control device for implementing the active balancing control method of the lithium-ion battery pack, including a data acquisition module, a data processing module, a data analysis module, a path planning module, an energy transfer device, a state monitoring module, and a parameter update module. The data acquisition module is used to acquire the state of charge (SOC) and health status of each individual battery cell in real time. The data processing module is used to determine the balancing priority of each individual battery cell based on its health status, adjust the SOC based on the balancing priority, obtain the decision SOC, and determine the balancing center point. The data analysis module is used to determine a correction factor based on the charge-discharge characteristic curve, determine the upper and lower limits of the balancing threshold based on the balancing center point, and mark individual batteries with a decision SOC higher than the upper limit of the balancing threshold as a discharge group and those lower than the lower limit of the balancing threshold as a charging group. The path planning module is used to construct an energy transfer demand matrix between the discharge group and the charging group, and generate an energy transfer path matrix. The energy transfer device is used to transfer energy from the individual batteries in the discharge group to the individual batteries in the charging group according to the energy transfer path matrix. The status monitoring module calculates the current state of charge (SOC) range of each individual battery cell, determines whether to end the equalization process, and checks if the number of equalization executions has reached the refresh count limit, deciding whether to recalculate the equalization priority and update the SOC threshold and refresh count limit. The parameter update module updates the SOC threshold and refresh count limit.

[0043] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An active balancing control method for a lithium-ion battery pack, characterized in that, Includes the following steps: Step 1: Obtain the state of charge and health status of each individual cell in the battery pack, determine the balancing priority of each individual cell based on the health status, and set the initial range threshold and refresh count limit. Step 2: Determine the correction factor based on the charge-discharge characteristic curve of the battery pack, adjust the state of charge according to the balancing priority to obtain the decision state of charge, and determine the balancing center point based on the decision state of charge. Step 3: Construct the energy transfer path matrix between individual cells by combining the correction factor and the equilibrium center point, and control the energy transfer device to perform energy transfer; Step 4: Calculate the range of the current state of charge of each individual cell. If the range is less than the range threshold, end the task; otherwise, proceed to step 8. Step 5: If the number of balance executions reaches the refresh count limit, update the refresh count limit and the range threshold, reset the balance execution count to zero, and return to Step 1; otherwise, increment the balance execution count by 1 and proceed to Step 2. The upper limit for the number of updates and the range threshold include: reading the maximum value of the decline in the health status of a single battery cell and the average decay rate of the state of charge range; calculating the aging factor based on the maximum value and the average decay rate; calculating the convergence factor based on the current number of equalization executions; and updating the upper limit for the number of updates and the range threshold based on the aging factor and the convergence factor.

2. The active balancing control method for lithium-ion battery packs according to claim 1, characterized in that, In step 1, determining the balancing priority of each individual battery cell based on its health status includes: determining the first priority of each individual battery cell as an energy outputter and the second priority as an energy receiver.

3. The active balancing control method for lithium-ion battery packs according to claim 1, characterized in that, In step 2, the correction factor is determined based on the position of the battery pack's state of charge (SOC) on the charge-discharge characteristic curve: during charging, when the SOC is in the initial polarization region or the final polarization region, the correction factor takes a fixed negative value; when the SOC is in the stable region, the correction factor changes linearly with the SOC. During discharging, when the SOC is in the initial polarization region or the final polarization region, the correction factor takes a fixed positive value; when the SOC is in the stable region, the correction factor changes linearly with the SOC.

4. The active balancing control method for lithium-ion battery packs according to claim 1, characterized in that, In step 2, the decision states of charge are sorted by numerical value, and the decision states of charge corresponding to the median are used as the initial center point. The sum of the absolute deviations between the decision states of charge of each individual cell and the initial center point is calculated. The initial center point is shifted in the direction with higher decision states of charge density. After each shift, the sum of absolute deviations is recalculated until the sum of absolute deviations no longer decreases. The center point at this time is determined as the equilibrium center point.

5. The active balancing control method for lithium-ion battery packs according to claim 1, characterized in that, In step 3, the upper limit of the equilibrium threshold Q1 = Q0 + [(1+f)δ] is calculated, and the lower limit of the equilibrium threshold Q2 = Q0 - [(1+f)δ] is calculated, where f is the correction factor, Q0 is the equilibrium center point, and δ is the basic equilibrium threshold. Individual cells with a decision state of charge higher than the upper limit of the equilibrium threshold are marked as discharge groups, and individual cells with a state of charge lower than the lower limit of the equilibrium threshold are marked as charging groups.

6. The active balancing control method for lithium-ion battery packs according to claim 5, characterized in that, In step 3, an energy transfer demand matrix is ​​constructed with the output energy of each individual battery in the discharge group as the row constraint and the demand energy of each individual battery in the charging group as the column constraint. The output energy is determined based on the difference between the current state of charge of the individual battery in the discharge group and the lower limit of the equilibrium threshold, and the demand energy is determined based on the difference between the upper limit of the equilibrium threshold and the current state of charge of the individual battery in the charging group.

7. The active balancing control method for lithium-ion battery packs according to claim 6, characterized in that, In step 3, with the goal of maximizing the sparsity of the energy transfer demand matrix, the single cell with the largest output energy in the discharge group is preferentially matched to the single cell with the largest energy demand in the charging group. This process is iterated until all energy demands are met or all output energy is exhausted. An energy transfer path matrix is ​​generated based on the matching results. The non-zero elements in this energy transfer path matrix represent the amount of energy transferred between the single cell in the discharge group and the single cell in the charging group.

8. The active balancing control method for lithium-ion battery packs according to claim 1, characterized in that, In step 3, the energy transfer device is any one of an isolated DC-DC converter, a non-isolated DC-DC converter, or an inductive equalizer.

9. The active balancing control method for lithium-ion battery packs according to claim 1, characterized in that, In step 5, a basic range threshold is determined based on the aging factor between a preset initial range threshold and a maximum allowable range threshold. The basic range threshold is then adjusted based on the convergence factor to obtain an updated range threshold. The updated range threshold is not lower than a preset protection threshold.

10. A control device for implementing the active balancing control method for a lithium-ion battery pack as described in claim 1, characterized in that, include: The data acquisition module is used to acquire the state of charge and health status of each individual battery cell in real time. The data processing module is used to determine the balancing priority of each individual cell based on its health status, adjust the state of charge based on the balancing priority, obtain the decision state of charge, and determine the balancing center point. The data analysis module is used to determine the correction factor based on the charge and discharge characteristic curves, determine the upper and lower limits of the balance threshold in combination with the balance center point, and mark the individual cells with a decision state of charge higher than the upper limit of the balance threshold as the discharge group and those with a state of charge lower than the lower limit of the balance threshold as the charging group. The path planning module is used to construct the energy transfer demand matrix between the discharge group and the charging group, and generate the energy transfer path matrix. An energy transfer device for transferring energy from individual cells in a discharge group to individual cells in a charging group according to the energy transfer path matrix; The status monitoring module is used to calculate the current range of the state of charge of each individual battery cell, determine whether to end the equalization process, and determine whether the number of equalization executions has reached the refresh limit, and decide whether to recalculate the equalization priority and update the range threshold and refresh limit. The parameter update module is used to update the range threshold and the upper limit of refresh count.

Images