A battery active equalization control method and system fusing health state and thermal risk barrier

Abstract

The application provides a battery active balancing control method and system fusing health state and thermal risk barrier, relates to the technical field of battery management system, and the control method comprises calculating the health correction available energy and the thermal risk index of each battery cell, so that the balancing target not only considers the difference of the state of charge, but also comprehensively corrects the influence of the health state and the temperature on the real available energy, thereby significantly improving the accuracy and energy utilization rate of the balancing control, and converting the safety indexes such as thermal risk and life damage into strict linear constraint conditions, and combining the candidate energy transfer path determined according to the energy transfer priority matrix, the optimal balancing current of each path is obtained by solving the optimization problem with linear constraints, while ensuring that the battery cell operates in the safe range, the path with high energy transfer benefit and low risk is preferentially selected for accurate regulation and control. The active balancing control system is used for realizing the battery active balancing control method.

Description

Technical Field

[0001] This invention relates to the field of battery management system technology, and specifically to a battery active balancing control method and system that integrates health status and thermal risk barrier. Background Technology

[0002] Lithium-ion battery packs are typically composed of multiple individual cells connected in series or in parallel. Due to differences in manufacturing processes, heat dissipation environments, connection internal resistance, and charge / discharge history, inconsistencies gradually emerge in the state of charge, capacity, internal resistance, temperature, and aging rate of each cell. When these inconsistencies continue to widen, some cells may reach the charging or discharging cutoff voltage prematurely, thereby reducing the usable capacity of the battery pack and increasing the risk of localized overheating and thermal runaway.

[0003] In existing technologies, 1. Existing active balancing technologies can transfer energy between battery cells through inductors, capacitors, transformers, or bidirectional DC / DC converters. Compared with passive balancing, they have advantages such as high energy utilization and low heat generation. However, many active balancing strategies still use voltage difference or SOC (remaining charge) difference as the main criteria, which easily ignores the differences in cell aging and thermal risk. For example, a battery cell with significantly degraded capacity and increased internal resistance may not be suitable for bearing a high balancing current even if its SOC is close to that of other battery cells. On the other hand, even if another battery cell has a high SOC, it should not continue to be used as an energy output side if its temperature is close to the safety limit. 2. Existing technologies also include schemes that incorporate temperature into the balancing control target, but only treat temperature deviation as a soft penalty term in the cost function. When the system faces rapid temperature rise or sudden increase in local internal resistance, the optimization solution may still give a large balancing current in a short period of time, causing the temperature or life damage index to exceed the safety boundary.

[0004] Therefore, there is an urgent need for an active equilibrium method that simultaneously considers available energy, health status, thermal risk, and hard safety constraints in equilibrium control. Summary of the Invention

[0005] This application provides a battery active balancing control method and system that integrates health status and thermal risk barrier to solve the above-mentioned technical problems.

[0006] In a first aspect, this application provides a battery active balancing control method that integrates health status and thermal risk barrier, characterized by comprising: In response to the battery pack meeting the equalization trigger condition, the state parameters of the cells in the battery pack are obtained. Determine the available energy for health correction of each cell based on the cell status parameters; Determine the thermal risk index of each battery cell based on its condition parameters; Construct a cost function that includes a deviation term for health-corrected available energy, a lifetime damage deviation term, a thermal risk penalty term, and an equilibrium loss term; Based on the upper limits of temperature, temperature rise rate, thermal risk, and lifespan damage, a control barrier function constraint is constructed, and the cost function and control barrier function constraint are transformed into an optimization problem with linear constraints. Candidate energy transfer paths are determined based on the energy transfer priority matrix, and with the candidate energy transfer paths fixed, the optimization problem with linear constraints is solved to obtain the optimal equilibrium current corresponding to each candidate path. Energy transfer is performed by controlling the bidirectional equalization converter based on the energy transfer path and the optimal equalization current.

[0007] Preferably, the cell status parameters include at least one of the following: state of charge, temperature, voltage, current, internal resistance, available capacity, and health status.

[0008] Preferably, the available energy for health correction of each cell is determined based on the state of charge, available capacity, and open-circuit voltage of each cell. ;

[0009] in, Indicates the first i Each battery cell at any time k The state of charge, Indicates the first i Each battery cell at any time k Available capacity Indicates the first i Each battery cell at any time k Open circuit voltage, This represents the energy availability coefficient, obtained by adjusting for health status and temperature.

[0010] Preferably, the energy availability factor It is determined by health status, internal resistance, and temperature, and the energy availability coefficient decreases as health status decreases, internal resistance increases, or temperature increases.

[0011] Preferably, the energy availability factor As shown in the following formula:

[0012] in, Indicates the first i Each battery cell at any time k ; health status; Indicates the first i Each battery cell at any time k The internal resistance; Indicates the reference internal resistance; Indicates the first i Each battery cell at any time k Temperature; Indicates reference temperature; and These represent the internal resistance correction factor and the temperature correction factor, respectively. express ; Indicates will Limited to Between 1 and 1; This indicates the lower limit of the energy availability coefficient.

[0013] Preferably, the thermal risk index is determined based on one or more of the following: cell temperature, temperature rise rate, internal resistance, predicted equalization current, and health status.

[0014] Preferably, the thermal risk index is determined based on at least one of the following quantities: i The temperature difference between the cell and the ambient temperature, the rate of temperature rise, the product of the square of the equalization current and the internal resistance, the increase of the internal resistance relative to the reference internal resistance, the amount of health state decay, and the extent to which the state of charge deviates from the safe state of charge range.

[0015] Preferred, the first i Each battery cell at any time k Output side thermal risk index As shown in the following formula:

[0016] in, Represents the normalization function. , to All represent weighting coefficients. Indicates the first i Each battery cell at any time k Temperature; Indicates ambient temperature; Indicates the rate of temperature rise; Indicates the first i Each battery cell at any time k The internal resistance; Indicates the reference internal resistance; Indicates the first i Each battery cell at any time k ; health status; Indicates the first i The degree to which each cell deviates from the safe SOC range.

[0017] Preferred, constructed cost function Satisfy the following formula:

[0018] in, Indicates battery cell i The energy available for health correction This indicates the available energy target value for health correction. Indicates the first i The amount of lifespan damage per battery cell Indicates the reference value for lifespan impairment. Indicates the first i Thermal risk index of the output side of each battery cell Indicates the first i The balanced loss of each battery cell Indicates the first i The equalization current of each cell This represents the equalization current of the previous control cycle. 、 、 、 and These represent the weighting coefficients for the available energy deviation for health correction, lifespan damage deviation, thermal risk, equilibrium loss, and current smoothing term, respectively.

[0019] Preferably, the lifespan damage amount is shown in the following formula:

[0020]

[0021] in, Indicates the first The increase in lifespan damage of each cell within the current control cycle. This indicates the amount of capacity loss in the current cycle. Indicates the rated capacity.

[0022] Preferably, the control barrier function constraints include temperature barrier function constraints and thermal risk barrier function constraints, and make... , Represents any barrier function constraint; Temperature barrier function constraint: , Indicates the first i The safety margin of each battery cell in terms of temperature. Indicates the upper limit of temperature. Indicates the first i Each battery cell at any time k Temperature; Thermal risk barrier function constraint: , Indicates the first i Safety margin for each battery cell in terms of thermal risk This indicates the upper limit of the heat risk index. The heat risk index is represented and normalized.

[0023] Preferably, when constructing an optimization problem with linear constraints, a first-order Taylor expansion is performed on the temperature barrier function constraint and the thermal risk barrier function constraint at the current operating point, which are then transformed into linear inequality constraints with respect to the equilibrium current.

[0024] Preferably, an energy transfer priority matrix is ​​constructed based on the health-corrected available energy, thermal risk index, health status, and converter losses of each battery cell. , Indicates from the first i The first cell to the first j Priority of energy transfer per battery cell; when The value is greater than the preset path opening threshold, and the first i The first cell and the first j When all cells meet the control barrier function constraints, the path is included in the candidate energy transfer path. If there are multiple candidate paths, the paths are sorted according to the priority matrix, and only the path with the highest priority is selected as the effective energy transfer path for the current control cycle. The remaining candidate paths are not executed for the time being and are recalculated in the next control cycle. Given a fixed candidate energy transfer path, the cost function, current upper and lower limits, converter power constraints, and linearized control barrier function constraints are combined to form an optimization problem with linear constraints. Solving this optimization problem yields the optimal equilibrium current corresponding to the candidate path.

[0025] Preferably, the energy transfer priority matrix is ​​as follows:

[0026] In the formula, The larger the value, the more significant the difference between the first and second digits. i The first cell to the first j The higher the priority of energy transfer from each battery cell; Indicates when the first i The energy available after cell health correction is higher than that of the first. j A positive transfer tendency only forms when each battery cell is used; This indicates the thermal risk index of the output-side battery cell; This indicates the amount of battery cell lifespan damage on the output side; Indicates from the first i The first battery cell to the first j Path loss of each battery cell | , Indicates battery cell To the battery cell The equalization current, Indicates path transmission efficiency. Indicates the control sampling period; This indicates a poor health status between the output and input sides; 、 、 、 and All of these represent path selection weight coefficients.

[0027] Preferably, the optimal equilibrium current is obtained by solving this optimization problem as shown in the following equation:

[0028]

[0029] in, This represents the optimization variable vector composed of all equalization currents in the prediction time domain; Represents a quadratic term matrix; Represents the coefficient vector of the linear term; and Let represent the inequality constraint matrix and the inequality constraint vector, respectively. and These represent the lower and upper limits of the equalization current, respectively. and Let represent the equality constraint matrix and equality constraint vector, respectively.

[0030] Preferably, after energy transfer, the internal resistance is estimated based on the balanced voltage, current, and temperature responses using recursive least squares, extended Kalman filtering, or a neural network. and available capacity Calculate the health status based on the capacity estimation results. The updated parameters are then fed back to the next control cycle to reacquire the status parameters of each cell.

[0031] Secondly, this application provides a battery active balancing control system that integrates health status and thermal risk barrier, used to implement a battery active balancing control method. The battery active balancing control system includes: The data acquisition module has its output terminals connected to the available energy calculation module, thermal risk calculation module, lifetime damage calculation module, and barrier function constraint construction module, respectively, and is used to send the acquired cell state parameters to each module. The available energy calculation module, whose output is connected to the input of the path selection module and the optimization solution module, is used to calculate the health correction available energy of each cell; The thermal risk calculation module, whose output is connected to the input of the barrier function constraint construction module, path selection module and optimization solution module, is used to calculate the thermal risk index of each cell. The lifetime damage calculation module has its output connected to the input of the barrier function constraint construction module, path selection module, and optimization solution module, and is used to calculate the lifetime damage of each cell. The barrier function constraint construction module has its output connected to the input of the optimization solution module and is used to construct control barrier function constraints. The path selection module, whose output is connected to the input of the optimization solution module, is used to generate candidate energy transfer paths; The optimization solution module, whose output is connected to the input of the equalization execution module, is used to construct and solve an optimization problem with linear constraints in order to obtain the optimal equalization current. The equalization execution module, whose output is electrically connected to the bidirectional equalization converter, is used to control the bidirectional equalization converter to perform energy transfer. The health status update module has its input connected to the output of the equalization execution module and its output connected to the input of the data acquisition module. It is used to feed back the updated parameters to the next control cycle.

[0032] The battery active balancing control method and system of this application have at least the following beneficial effects: The control method of this application determines the health-corrected available energy and thermal risk index of each cell based on the cell state parameters. This ensures that the balancing objective not only considers the differences in state of charge but also comprehensively corrects the impact of health status and temperature on the actual available energy, thereby significantly improving the accuracy of balancing control and energy utilization. Simultaneously, by constructing a cost function that includes deviation terms for health-corrected available energy, lifetime damage deviation terms, thermal risk penalty terms, and balancing loss terms, energy balancing, lifetime management, thermal risk control, and energy consumption management are all integrated into the optimization framework, achieving multi-objective synergistic optimization. Furthermore, safety indicators such as thermal risk and lifetime damage are transformed into strict linear constraints. Combined with candidate energy transfer paths determined based on the energy transfer priority matrix, the linearly constrained optimization problem is solved under the premise of fixed paths to obtain the optimal balancing current for each path. While ensuring that the cells operate within a safe range, the method prioritizes paths with high energy transfer efficiency and low risk for precise control, balancing balancing efficiency and system safety. This solves the problems of single control objectives, insufficient safety, and low energy efficiency in existing balancing methods. Attached Figure Description

[0033] Figure 1 This is a flowchart of Embodiment 1 of this application; Figure 2 This is a schematic diagram of the control barrier function constraint construction in Example 1; Figure 3 This is a flowchart of Embodiment 2 of this application; Figure 4 This is a schematic diagram of the control system according to Embodiment 3 of this application; Figure 5 This is a schematic diagram of the energy transfer path of the bidirectional equalizer in Embodiment 3; Figure 6 Figure 5 Diagram of bidirectional Cuke equalization circuit for bidirectional equalizer converter. Detailed Implementation

[0034] The present invention will be further described below with reference to embodiments. It should be understood that the described embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0035] Example 1: This embodiment discloses a battery active balancing control method that integrates health status and thermal risk barriers, to address the problems of existing active balancing strategies that easily ignore differences in cell health status, cannot impose hard constraints on thermal risks, and may over-balance aging cells. The battery in this embodiment includes a lithium-ion battery pack.

[0036] The battery active balancing control method in this embodiment includes the following steps, such as: Figure 1 As shown: Step S100: In response to the battery pack meeting the equalization trigger condition, obtain the cell state parameters in the battery pack; The equalization triggering conditions may include a health correction available energy difference greater than a first threshold, a thermal risk index difference greater than a second threshold, a SOC difference greater than a third threshold, or the battery management system being in maintenance equalization mode. The first to third thresholds are set according to actual needs. It should be noted that the specific content of the equalization triggering conditions is not the core of this embodiment; those skilled in the art can choose under what conditions to actively or passively trigger the equalization based on actual needs.

[0037] In this embodiment, the cell status parameters include state of charge, temperature, voltage, current, internal resistance, available capacity, and health status. In some other embodiments, the cell status parameters may also include cell thermal capacity parameters, etc.

[0038] Step S200: Determine the health correction available energy for each cell based on its state of charge, available capacity, and open-circuit voltage. Unlike directly using SOC, this embodiment uses available capacity, open circuit voltage, and health status correction coefficient to jointly determine the energy that the cell can contribute to the outside world in the current state (that is, the amount of electricity that can be output).

[0039]

[0040] in, Indicates the first i Each battery cell at any time k The state of charge, Indicates the first i Each battery cell at any time k Available capacity Indicates the first i Each battery cell at any time k Open circuit voltage, This represents the energy availability coefficient.

[0041] In this embodiment, the balancing target is changed from consistent SOC to consistent available energy after health correction, which is more in line with the actual available capacity requirements of battery packs with capacity degradation.

[0042] Furthermore, the energy availability factor The effect of aging, internal resistance, and thermal safety margin on available energy can be obtained through health status and temperature corrections.

[0043]

[0044] in, Indicates the first i Each battery cell at any time k The health status is obtained through the BMS system; Indicates the first i Each battery cell at any time k The internal resistance; Indicates the reference internal resistance; Indicates the first i Each battery cell at any time k Temperature; Indicates reference temperature; and These represent the internal resistance correction factor and the temperature correction factor, respectively. express ; Indicates will Limited to Between 1 and 1; This indicates the lower limit of the energy availability factor (the specific value can be set according to actual needs).

[0045] As can be seen from the above formula, when the SOH of the battery cell decreases, the internal resistance increases, or the temperature is higher than the reference temperature, the effective energy that can participate in the balancing is reduced accordingly.

[0046] Furthermore, the reference internal resistance in this embodiment Reference temperature Internal resistance correction factor and temperature correction factor It can be obtained based on existing calibration or correction methods, as follows: Reference internal resistance The battery cells are standard new cells, operating at a reference temperature of 25℃ and a midpoint of 50% SOC. The DC internal resistance of the cells is measured using the HPPC pulse internal resistance test method. This value is directly used as the unified reference internal resistance for this batch of cells. The reference internal resistance of the cascaded batteries is calibrated separately for each batch.

[0047] Reference temperature The industry standard test temperature for lithium batteries is 25℃, which is fixed for all batches. Only in extreme high-temperature energy storage scenarios can it be modified offline to 30℃, which is a fixed reference value for the industry standard.

[0048] Internal resistance correction factor At a constant temperature of 25℃, multiple sets of different DC internal resistances were applied to the battery cell, and multiple sets of data were collected. Fit an exponential decay function to the actual usable energy decay data of the battery cell. The least squares fitting was used to obtain ; Temperature correction factor With fixed internal resistance and constant SOC, multiple stepped ambient temperatures are set above [the desired temperature range]. The relationship between temperature and available energy reduction was collected, and an exponential curve was fitted. Least squares solution .

[0049] It should be noted that HPPC pulse internal resistance testing, temperature-capacity decay fitting, and least squares parameter fitting are all conventional and mature technologies for offline calibration of power battery BMS. Furthermore, lithium iron phosphate batteries have low internal resistance sensitivity, which can reduce... Reduce internal resistance; ternary lithium batteries are thermally sensitive, which can increase... Strengthen high-temperature restrictions.

[0050] Step S300: Determine the thermal risk index of each cell based on its temperature, temperature rise rate, internal resistance, predicted equalization current, and health status. The predicted equalization current is taken from the optimal equalization current of the previous control cycle. If it is the initial control cycle, the predicted equalization current is set to zero or designed to be a very small value.

[0051] Furthermore, the thermal risk index in this embodiment comprehensively considers cell temperature, temperature rise rate, internal resistance heat generation, health status decay, and deviation from the SOC safe range to measure the likelihood of the cell entering a high thermal risk state during the equilibration process. Preferably, when the thermal risk index is greater than or equal to a preset thermal risk upper limit, the cell is determined to be in a high thermal risk state. i The individual battery cell is in a state of high heat risk.

[0052]

[0053] in, Indicates the first i Each battery cell at any time k The thermal risk index on the output side Represents the normalization function. , to All represent weighting coefficients. Indicates the first i Each battery cell at any time k Temperature; Indicates ambient temperature; Indicates the rate of temperature rise; Indicates the first i Each battery cell at any time k The internal resistance; Indicates the reference internal resistance; Indicates the first i Each battery cell at any time k health status ； Indicates the first i The degree to which each cell deviates from the safe SOC range.

[0054] In this embodiment, the thermal risk index is composed of multiple different physical quantities such as temperature, rate of temperature rise, internal resistance, and health status. Direct weighting would result in inconsistent dimensions. Therefore, a normalization process is added when calculating the thermal risk index to convert each risk factor into a dimensionless form before weighting and fusion, so that the final thermal risk index maintains dimensionless characteristics and can be used for thermal risk threshold comparison and barrier function constraints. Furthermore, weighting coefficients to The system employs least squares regression training with a multi-condition thermal risk dataset for power batteries. This data labeling and linear regression fitting training method is a well-known technique in the field of battery thermal state assessment. After training, normalization yields fixed weights suitable for vehicle or energy storage scenarios, and offline retraining and adjustment are possible based on the device's heat dissipation level. This embodiment provides a set of example values: (Temperature difference between battery cell and environment) (For the rate of temperature rise, thermal safety has the highest priority.) (Internal resistance heat generation) (Internal resistance increase term) (SOH attenuation term) (SOC safety range deviation).

[0055] Furthermore,

[0056] In the above formula, and These are the lower and upper limits of the safe SOC range, respectively; when When within the safe zone, ;when When it is below the lower limit or above the upper limit, As the degree of deviation increases.

[0057] Step S400: Construct a multi-objective cost function The cost function simultaneously considers health correction available energy consistency, lifetime damage consistency, thermal risk penalty, equilibrium loss and current smoothness.

[0058]

[0059] Preferably, in this embodiment, multi-step predictive control is adopted to further improve stability, and the cost function is... As shown below:

[0060] in, Indicates the number of battery cells. To predict the length of the time domain, Indicates the prediction step number. Indicates battery cell i At any moment The energy available for health correction This indicates the available energy target value for health correction. Indicates the first i Each battery cell at any time The amount of lifespan damage, Indicates the reference value for lifespan impairment. Indicates the first i Each battery cell at any time The thermal risk index on the output side Indicates the first i Each battery cell at any time Balanced loss, Indicates the first i Each battery cell at any time The equalization current, , , , and These represent the weighting coefficients for health correction available energy deviation, lifetime damage deviation, thermal risk, equalization loss, and current smoothing term, respectively. Further preferred values ​​for these weighting coefficients are: , , , , .

[0061] Furthermore, in the aforementioned cost function In the middle, lifespan damage Used for the accumulation equalization process and normal charge / discharge process for the first i The relative aging effect caused by individual battery cells.

[0062]

[0063] In the formula, Indicates the first i Each battery cell at any time k The +1 lifetime damage amount can be used as the input to the lifetime damage barrier and cost function for the next control cycle. Indicates the first i Each battery cell at any time k The amount of lifespan damage, that is, the amount of lifespan damage within the current control cycle; Indicates the first The incremental lifespan damage of each cell within the current control cycle.

[0064] The lifetime damage increment function is obtained using the battery capacity decay model, defining the first... The incremental lifespan damage of each cell within the current control cycle is:

[0065] in, This indicates the amount of capacity loss in the current cycle. This indicates the rated capacity. The capacity loss can be obtained from an experimentally calibrated battery aging model, for example: ,in This represents the aging rate coefficient obtained through cycle life testing.

[0066] Step S500, as follows Figure 2 As shown, step S500 includes steps S510 and S520. Step S510 is used to construct control barrier function constraints, and step S520 is used to transform the cost function and control barrier function constraints into an optimization problem with linear constraints.

[0067] Step S510: Construct control barrier function constraints. In this embodiment, the upper limit of temperature and the upper limit of thermal risk are defined as the safety set boundaries, respectively.

[0068] Temperature barrier function constraint: , Indicates the first i The safety margin of each cell in terms of temperature, and at the same time, the constraint condition of the temperature barrier function; Indicates the upper limit of temperature. Indicates the first i Each battery cell at any time k Temperature; Thermal risk barrier function constraint: , Indicates the first i Safety margin for each battery cell in terms of thermal risk This indicates the upper limit of the heat risk index. This indicates the heat risk index.

[0069] Preferably, the above barrier function constraints satisfy: , It can represent or Any one of the following; this constraint is used to ensure that the safety margin at the next time step does not decrease too quickly, so that the state remains in the safety set in the prediction time domain.

[0070] Step S520: Transform into constraints in the optimization solution.

[0071] Because there is a nonlinear relationship between the barrier function constraints (temperature barrier function constraints or thermal risk barrier function constraints) and the equilibrium current, it is impossible to directly form linear constraints suitable for rolling optimization solutions. Therefore, in the first... Within each control cycle, a reference point for equalization current linearization is selected. The previous cycle equilibrium current is preferred as the current reference point, and a first-order Taylor expansion of the nonlinear barrier function constraint is performed near this point:

[0072] in: Indicates the first Individual cell regarding equalization current Barrier function constraints; Indicates the first The balanced current to be optimized in each control cycle; Indicates the first The linearized reference equalization current for each control cycle can be determined by the output value of the previous control cycle or the current predicted equalization current. This represents the gradient of the barrier function with respect to the equilibrium current at the reference point.

[0073] Based on the above linear approximation, the original nonlinear barrier function is constrained as follows:

[0074] Transform this into a linear inequality constraint concerning the equilibrium current:

[0075] in:

[0076] in, Indicates the first Linearization coefficients for each control cycle Indicates the first Each control cycle corresponds to a constraint boundary.

[0077] Preferably, the lifetime damage increment is not related to the equalization current, so it can be excluded from linearization and treated as a term in the cost function.

[0078] Step S600 includes steps S610, S620 and S630, wherein step S610 includes determining the energy transfer path, step S620 includes solving the optimization problem with linear constraints, and step S630 includes calculating the optimal equilibrium current corresponding to each candidate path.

[0079] Step S610: Determine the energy transfer path. Based on the available energy for health correction, thermal risk index, health status, and converter losses of each cell, construct an energy transfer priority matrix. ,in Indicates from the first i The first cell to the first j Priority of energy transfer per battery cell;

[0080] In the formula, The larger the value, the more significant the difference between the first and second digits. i The first cell to the first j The higher the priority of energy transfer from each battery cell; Indicates when the first i The energy available after cell health correction is higher than that of the first. j A positive transfer tendency only forms when each battery cell is used; This indicates the thermal risk index of the output-side battery cell; This indicates the amount of battery cell lifespan damage on the output side; Indicates from the first i The first battery cell to the first j Path loss of each battery cell | ， Indicates battery cell To the battery cell The equalization current, This represents the path transmission efficiency (the transmission efficiency is taken as 0.95 based on simulation results). Indicates the control sampling period; This indicates a poor health status between the output and input sides; 、 、 、 and All represent path selection weighting coefficients, used to adjust the impact of energy difference, thermal risk, lifetime damage, path loss, and poor health status on path priority. In practice, thermal safety takes priority, and further optimization is performed. , , , , .

[0081] when Greater than the path opening threshold and the first i The first battery cell, the first j When all cells satisfy the barrier function constraint, the corresponding path is included in the candidate energy transfer path. This mechanism can avoid outputting energy from cells with high heat risk or inputting energy into cells close to the upper temperature limit. Preferably, if there are multiple candidate paths, they are selected according to the priority matrix. The paths are sorted, and only the path with the highest priority is selected as the effective energy transfer path for the current control cycle. The other candidate paths are not executed for the time being and are recalculated in the next control cycle. The path opening threshold can be set according to the actual operating conditions. Step S620: With the candidate energy transfer path fixed, the cost function is combined with the current upper and lower limit constraints, the converter power constraints, and the linearized control barrier function constraints to form an optimization problem with linear constraints. The optimization problem is then solved. The optimization problem can be a quadratic programming problem. When it is necessary to select the discrete path switching state at the same time, the upper-level path selection module can first determine the candidate path, and then the lower-level continuous optimization module can solve the equalization current.

[0082] Step S630: Calculate the optimal equalization current for each candidate path, as shown in the following formula:

[0083]

[0084] in, This represents the optimization variable vector composed of all equalization currents in the prediction time domain; This represents a quadratic term matrix, which is usually a positive definite or positive semi-definite matrix. Represents the coefficient vector of the linear term; and Let represent the inequality constraint matrix and the inequality constraint vector, respectively. Together, they constitute a set of linear inequality constraints concerning the equilibrium current. This set of inequality constraints includes the upper and lower limits of the equilibrium current, the converter power limit, and the linearized control barrier function constraints. and These represent the lower and upper limits of the equalization current, respectively. and These represent the equality constraint matrix and the equality constraint vector, respectively. The constraints formed by their combination include energy conservation constraints or multi-port converter topology constraints.

[0085] Step S700: Control the bidirectional equalizer according to the solution result. The bidirectional equalizer is configured as a multi-port bidirectional equalizer. In this embodiment, the bidirectional equalizer includes multiple bidirectional Cuke units. The bidirectional equalizer can also use other bidirectional Buck-Boost units, isolated bidirectional DC / DC units, or combinations thereof.

[0086] The optimal balanced current is used to generate the duty cycle, phase shift angle, or PWM drive signal of the switching transistor, thereby realizing energy transfer from cell to cell, from cell to bus, or from bus to cell.

[0087] Duty cycle

[0088] in, This represents the optimal equilibrium current. Indicates balanced inductance. Indicates the switching frequency. Indicates battery cell The voltage.

[0089] Example 2: The difference between this embodiment two and embodiment one is that this embodiment two further includes step S800 after step S700. Step S800 includes online updating of the internal resistance, available capacity, health status, and lifetime damage of each cell, such as... Figure 3 As shown, the details are as follows: Based on the balanced voltage, current, and temperature data, the internal resistance and usable capacity of each cell are identified online using recursive least squares method, extended Kalman filter, or neural network model. ; The state of health (SOH) is calculated based on the updated available capacity. , Indicates the rated capacity; Calculate the incremental lifespan damage during this control cycle based on the cell temperature, equalization current, state of charge (SOC), and state of harmonics (SOH). and according to Recursively update lifetime damage; The updated internal resistance, available capacity, health status, and lifetime damage are fed back to the next control cycle as initial values ​​after reacquiring the state parameters of each cell. These values ​​are then used for health correction, available energy calculation, thermal risk index calculation, control barrier function constraint construction, and cost function calculation in the next control cycle until an equilibrium state is reached.

[0090] Example 3: This embodiment discloses a battery active balancing control system that integrates health status and thermal risk barriers, used to implement the battery active balancing control method in Embodiment 1 or Embodiment 2. The battery active balancing control system includes a data acquisition module, an available energy calculation module, a thermal risk calculation module, a lifetime damage calculation module, a barrier function constraint construction module, a path selection module, an optimization solution module, and a balancing execution module. More preferably, the battery active balancing control system also includes a health status update module. These modules are all part of the BMS main control unit, such as... Figure 4 As shown: The data acquisition module is used to send the acquired cell status parameters to each module; The available energy calculation module is used to calculate the available energy for health correction of each cell; The thermal risk calculation module is used to calculate the thermal risk index of each battery cell; The life damage calculation module is used to calculate the life damage of each cell; The barrier function constraint building module is used to construct control barrier function constraints; The path selection module is used to generate candidate energy transfer paths; The optimization solution module is used to construct and solve optimization problems with linear constraints and calculate the optimal equilibrium current; the optimization solution module can be deployed in the BMS main control unit, domain controller or edge computing unit.

[0091] The equalization execution module is used to control the bidirectional equalization converter to perform energy transfer; The health status update module is used to feed back the updated parameters to the next control cycle. Preferably, the health status update module is connected to the memory to save the historical parameters of each cell.

[0092] The battery active balancing control system also includes a sensor array, a bidirectional balancing converter, a thermal management unit, and an edge / cloud parameter update unit.

[0093] The sensor array is used to acquire various relevant physical quantities of the battery pack (voltage, current, and temperature, etc.) and is connected to the data acquisition module.

[0094] The bidirectional equalization converter is connected to the equalization execution module. The equalization execution module generates a corresponding PWM drive signal (duty cycle control signal) based on the received current command. This signal is sent to the multi-port bidirectional equalization converter, which is connected to the battery pack through a bidirectional energy transfer channel, forming a physical energy flow loop in the system and realizing the redistribution of energy between the battery cells. Figure 5 As shown, Figure 5 In Indicates the first Each battery cell. The bidirectional equalization converter includes a bidirectional Cuke unit; see [link to bidirectional Cuke unit]. Figure 6 Specifically, this includes battery cells connected in sequence. First Inductor First capacitor Second inductor Second capacitor First switching transistor One end is connected between the first inductor and the first capacitor, and the other end is connected between the battery cell and the second capacitor. The second switching transistor... One end is connected between the first capacitor and the second inductor, and the other end is connected between the battery cell and the second capacitor, wherein the second capacitor and the other battery cell share energy.

[0095] The thermal management unit is connected to the thermal risk calculation module. When the thermal risk index calculation module assesses that the current battery pack has a high thermal risk (such as the temperature rising too quickly or local hot spots), it will send a thermal management request to the thermal management interface.

[0096] The edge / cloud parameter update unit is connected to the health status update module, which facilitates the uploading of relevant parameters to the edge / cloud parameter update unit.

[0097] Example 4: This fourth embodiment describes the working process of the battery active balancing control system in the third embodiment at various stages: The battery management system collects the terminal voltage, current, and temperature of N series-connected cells at a preset sampling period, and estimates the state of charge (SOC) based on the open-circuit voltage model, equivalent circuit model, or data-driven model. Simultaneously, the system estimates the internal resistance of each cell based on the voltage recovery curves and pulse current responses of the most recent charge / discharge cycles. and available capacity and calculate .

[0098] During the equalization triggering phase, the available energy for health correction of each cell is first calculated. ,like If the energy difference threshold is exceeded, or the difference in lifespan damage between any two cells exceeds the lifespan difference threshold, the cell enters the lifespan-aware balancing mode. Unlike traditional SOC balancing, in the lifespan-aware balancing mode, the reference target for the cell is not a fixed average SOC, but a target value for healthy correction of available energy. For cells with low SOH, high internal resistance, or small thermal safety margin, the upper limit of the current for participating in balancing is automatically reduced.

[0099] During the thermal risk calculation phase, the controller uses temperature, temperature rise rate, internal resistance heat generation, and SOH decay to calculate the thermal risk index. ,when When the derating threshold is exceeded, the system does not immediately exit the balancing process. Instead, it gradually moves the cell away from a high-risk state by lowering the balancing current limit and increasing the thermal risk weight. When the thermal risk limit is reached or exceeded, the system stops the energy transfer path associated with the battery cell and outputs a thermal safety alarm to the vehicle controller or energy storage management system.

[0100] During the barrier function constraint construction phase, the upper limit of temperature, the upper limit of risk, and the upper limit of lifetime damage are constructed into a safety set. For example, the temperature safety set is... Thermal risk safety set is In order to keep the state in the prediction time domain within a safe set, for , and A discrete-time control barrier function constraint is applied, which, after linearization, together with the equilibrium current constraint, serves as an inequality constraint for the optimization problem.

[0101] During the path selection phase, based on the energy transfer priority matrix M Generate a set of candidate paths, if the first path is... i The energy available for health correction of each cell is higher than that of the first. j The first cell, and the first i If the thermal risk index of a cell is low and its internal resistance is not higher than the benchmark threshold, then starting from the [number]th cell... i The first cell to the first j The first cell has higher priority for transferring energy, and vice versa. i If a cell's temperature is close to its upper limit or its state of equilibrium (SOH) is significantly low, it will not be prioritized as the energy output side, even if its state of equilibrium (SOC) is high.

[0102] In the continuous optimization phase, the candidate paths are fixed, and the optimal equalization current of each path in the prediction time domain is solved. The first output control quantity is applied to the multi-port bidirectional equalization converter. The data is re-acquired and the optimization is rolled over in the next sampling period. In this way, this embodiment has both the foresight of model predictive control and the use of barrier functions to ensure thermal safety boundaries.

[0103] For secondary battery packs, the differences in cell sources and historical operating conditions are greater, and the dispersion of state of harmonics (SOH) and internal resistance is more obvious. With the adoption of this embodiment, the system can automatically identify cells with high internal resistance or low SOH, reduce their proportion of balancing current, and allocate more balancing tasks to cells with better health and low loss paths, thereby delaying the expansion of inconsistency in secondary battery packs.

[0104] For battery clusters in energy storage power stations, the equalization control in this embodiment can be initiated during low-load periods or planned maintenance periods. The system can use the power plan of the energy storage converter, the status of the battery cluster temperature control system, and the ambient temperature as external constraint inputs. When the ambient temperature is high or the cooling capacity is insufficient, the thermal risk weight is automatically increased, making the equalization strategy more conservative; when the ambient temperature is suitable and the cooling capacity is sufficient, the system appropriately increases the upper limit of the equalization current to shorten the equalization time.

[0105] The modules in the above embodiments can be implemented by hardware circuits, by a processor executing a program stored in a computer-readable storage medium, or by a combination of hardware and software. All equivalent substitutions, modifications, and combinations made within the spirit and principles of this invention should fall within the protection scope of this invention.

Claims

1. A battery active balancing control method integrating health status and thermal risk barrier, characterized in that, include: In response to the battery pack meeting the equalization trigger condition, the state parameters of the cells in the battery pack are obtained. Determine the available energy for health correction of each cell based on the cell status parameters; Determine the thermal risk index of each battery cell based on its condition parameters; Construct a cost function that includes a bias term for health-corrected available energy, a lifetime damage bias term, a thermal risk penalty term, and an equilibrium loss term; Based on the upper limits of temperature, temperature rise rate, thermal risk, and lifespan damage, a control barrier function constraint is constructed, and the cost function and control barrier function constraint are transformed into an optimization problem with linear constraints. Candidate energy transfer paths are determined based on the energy transfer priority matrix, and with the candidate energy transfer paths fixed, the optimization problem with linear constraints is solved to obtain the optimal equilibrium current corresponding to each candidate path. Energy transfer is performed by controlling the bidirectional equalization converter based on the energy transfer path and the optimal equalization current.

2. The battery active balancing control method according to claim 1, characterized in that, Cell status parameters include at least one of the following: state of charge, temperature, voltage, current, internal resistance, available capacity, and health status.

3. The battery active balancing control method according to claim 2, characterized in that, The available energy for health correction of each cell is determined based on its state of charge, available capacity, and open-circuit voltage. ; in, Indicates the first i Each battery cell at any time k The state of charge, Indicates the first i Each battery cell at any time k Available capacity Indicates the first i Each battery cell at any time k Open circuit voltage, This represents the energy availability coefficient, obtained by adjusting for health status and temperature.

4. The battery active balancing control method according to claim 3, characterized in that, Energy availability factor It is determined by health status, internal resistance, and temperature, and the energy availability coefficient decreases as health status decreases, internal resistance increases, or temperature increases.

5. The control method according to claim 4, characterized in that, Energy availability factor As shown in the following formula: in, Indicates the first i Each battery cell at any time k ; health status; Indicates the first i Each battery cell at any time k The internal resistance; Indicates the reference internal resistance; Indicates the first i Each battery cell at any time k Temperature; Indicates reference temperature; and These represent the internal resistance correction factor and the temperature correction factor, respectively. express ; Indicates will Limited to Between 1 and 1; This indicates the lower limit of the energy availability coefficient.

6. The battery active balancing control method according to any one of claims 2 to 5, characterized in that, The thermal risk index is determined based on one or more of the following: cell temperature, rate of temperature rise, internal resistance, predicted equalization current, and health status.

7. The battery active balancing control method according to claim 6, characterized in that, The thermal risk index is determined based on at least one of the following quantities: i The temperature difference between the cell and the ambient temperature, the rate of temperature rise, the product of the square of the equalization current and the internal resistance, the increase of the internal resistance relative to the reference internal resistance, the amount of health state decay, and the extent to which the state of charge deviates from the safe state of charge range.

8. The battery active balancing control method according to claim 7, characterized in that, No. i Each battery cell at any time k Output side thermal risk index As shown in the following formula: in, Represents the normalization function. , to All represent weighting coefficients. Indicates the first i Each battery cell at any time k Temperature; Indicates ambient temperature; Indicates the rate of temperature rise; Indicates the first i Each battery cell at any time k The internal resistance; Indicates the reference internal resistance; Indicates the first i Each battery cell at any time k ; health status; Indicates the first i The degree to which each cell deviates from the safe SOC range.

9. The battery active balancing control method according to claim 1, characterized in that, Constructed cost function Satisfy the following formula: in, Indicates battery cell i The energy available for health correction This indicates the available energy target value for health correction. Indicates the first i The amount of lifespan damage per battery cell Indicates the reference value for lifespan impairment. Indicates the first i Thermal risk index of the output side of each battery cell Indicates the first i The balanced loss of each battery cell Indicates the first i The equalization current of each cell This represents the equalization current of the previous control cycle. 、 、 、 and These represent the weighting coefficients for the available energy deviation for health correction, lifespan damage deviation, thermal risk, equilibrium loss, and current smoothing term, respectively.

10. The battery active balancing control method according to claim 9, characterized in that, The lifespan damage is shown in the following formula: in, Indicates the first The increase in lifespan damage of each cell within the current control cycle. This indicates the amount of capacity loss in the current cycle. Indicates the rated capacity.

11. The battery active balancing control method according to claim 10, characterized in that, Control barrier function constraints include temperature barrier function constraints and thermal risk barrier function constraints, and enable... , Represents any barrier function constraint; Temperature barrier function constraint: , Indicates the first i The safety margin of each battery cell in terms of temperature. Indicates the upper limit of temperature. Indicates the first i Each battery cell at any time k Temperature; Thermal risk barrier function constraint: , Indicates the first i Safety margin for each battery cell in terms of thermal risk This indicates the upper limit of the heat risk index. The heat risk index is represented and normalized.

12. The battery active balancing control method according to claim 11, characterized in that, When constructing the optimization problem with linear constraints, a first-order Taylor expansion is performed on the temperature barrier function constraint and the thermal risk barrier function constraint at the current operating point, which is transformed into a linear inequality constraint on the equilibrium current.

13. The battery active balancing control method according to claim 12, characterized in that, Based on the health-corrected available energy, thermal risk index, health status, and converter losses of each battery cell, an energy transfer priority matrix is ​​constructed. , Indicates from the first i The first cell to the first j Priority of energy transfer per battery cell; when The value is greater than the preset path opening threshold, and the first i The first cell and the first j When all cells meet the control barrier function constraints, the path is included in the candidate energy transfer path. If there are multiple candidate paths, the paths are sorted according to the priority matrix, and only the path with the highest priority is selected as the effective energy transfer path for the current control cycle. The remaining candidate paths are not executed for the time being and are recalculated in the next control cycle. Given a fixed candidate energy transfer path, the cost function, current upper and lower limits, converter power constraints, and linearized control barrier function constraints are combined to form an optimization problem with linear constraints. Solving this optimization problem yields the optimal equilibrium current corresponding to the candidate path.

14. The battery active balancing control method according to claim 13, characterized in that, The energy transfer priority matrix is ​​shown in the following formula: In the formula, The larger the value, the more significant it is from the first... i The first cell to the first j The higher the priority of energy transfer from each battery cell; Indicates when the first i The energy available after cell health correction is higher than that of the first. j A positive transfer tendency only forms when each battery cell is used; This indicates the thermal risk index of the output-side battery cell; This indicates the amount of battery cell lifespan damage on the output side; Indicates from the first i The first battery cell to the first j Path loss of each battery cell | , Indicates battery cell To the battery cell The equalization current, Indicates path transmission efficiency. Indicates the control sampling period; This indicates a poor health status between the output and input sides; 、 、 、 and All of these represent path selection weight coefficients.

15. The battery active balancing control method according to claim 13 or 14, characterized in that, Solving this optimization problem yields the optimal equilibrium current as shown in the following equation: in, This represents the optimization variable vector composed of all equalization currents in the prediction time domain; Represents a quadratic term matrix; Represents the coefficient vector of the linear term; and Let represent the inequality constraint matrix and the inequality constraint vector, respectively. and These represent the lower and upper limits of the equalization current, respectively. and Let represent the equality constraint matrix and equality constraint vector, respectively.

16. The battery active balancing control method according to claim 15, characterized in that, After energy transfer, Based on the balanced voltage, current, and temperature responses, the internal resistance is estimated using recursive least squares, extended Kalman filtering, or a neural network. and available capacity Health status is calculated based on capacity estimation results. The updated parameters are then fed back to the next control cycle to reacquire the status parameters of each cell.

17. A battery active balancing control system integrating health status and thermal risk barrier, characterized in that, For implementing the battery active balancing control method according to any one of claims 1 to 16, the battery active balancing control system comprises: The data acquisition module has its output terminals connected to the available energy calculation module, thermal risk calculation module, lifetime damage calculation module, and barrier function constraint construction module, respectively, and is used to send the acquired cell state parameters to each module. The available energy calculation module, whose output is connected to the input of the path selection module and the optimization solution module, is used to calculate the health correction available energy of each cell; The thermal risk calculation module, whose output is connected to the input of the barrier function constraint construction module, path selection module and optimization solution module, is used to calculate the thermal risk index of each cell. The lifetime damage calculation module has its output connected to the input of the barrier function constraint construction module, path selection module, and optimization solution module, and is used to calculate the lifetime damage of each cell. The barrier function constraint construction module has its output connected to the input of the optimization solution module and is used to construct control barrier function constraints. The path selection module, whose output is connected to the input of the optimization solution module, is used to generate candidate energy transfer paths; The optimization solution module, whose output is connected to the input of the equalization execution module, is used to construct and solve an optimization problem with linear constraints in order to obtain the optimal equalization current. The equalization execution module, whose output is electrically connected to the bidirectional equalization converter, is used to control the bidirectional equalization converter to perform energy transfer. The health status update module has its input connected to the output of the equalization execution module and its output connected to the input of the data acquisition module. It is used to feed back the updated parameters to the next control cycle.

Images