Liquid cooling temperature control method for power battery

Abstract

The application discloses a liquid cooling temperature control method of a power battery, comprising the following steps: acquiring a dynamic thermal management demand index corresponding to a working condition mode; acquiring a thermal coupling model of the power battery; generating a target equivalent convective heat exchange coefficient of the power battery based on a feedforward heat exchange coefficient and a feedback heat exchange coefficient, and optimizing the thermal coupling model according to the target equivalent convective heat exchange coefficient; determining optimal control parameters of a liquid cooling system based on a preset multi-objective optimization algorithm and according to the dynamic thermal management demand index, the optimized thermal coupling model, an inlet temperature and a volume flow of cooling liquid; and controlling the liquid cooling system according to the optimal control parameters. Thus, the application solves the technical problem that the heat exchange coefficient does not match the real heat exchange capacity, realizes real-time and accurate tracking of the equivalent convective heat exchange coefficient, provides an accurate model basis for generation of a liquid cooling system control strategy, and ensures the accuracy of the heat exchange coefficient at the moment of flow state mutation.

Description

Technical Field

[0001] This application relates to the field of liquid cooling temperature control technology for power batteries, and in particular to a liquid cooling temperature control method for power batteries. Background Technology

[0002] In the field of power batteries for new energy vehicles, liquid cooling temperature control technology is the mainstream technical solution for power battery thermal management. In related technologies, battery liquid cooling control methods typically involve pre-training a control command model, which then outputs target control commands for the battery liquid cooling system. These target control commands are then used to control the battery liquid cooling system. However, the training and output of this control command model rely on pre-collected external information such as terrain data and driving behavior data. The target control commands generated by the model are essentially static mappings based on historical data and preset operating conditions. When the battery pack is operating under wide temperature range conditions, especially during the drastic transition from low-temperature to normal-temperature conditions, the viscosity of the coolant decreases exponentially with increasing temperature. This drastic change in viscosity causes a sudden shift between laminar and turbulent flow within the flow channels, resulting in a nonlinear jump in the actual heat transfer coefficient. Existing control command models cannot detect changes in heat transfer capacity caused by sudden changes in coolant properties. The control commands output by the model are based on open-loop control with a fixed heat transfer coefficient. When the flow regime undergoes a sudden change across the critical Reynolds number, the flow rate and temperature commands calculated by the controller based on the hysteresis heat transfer coefficient are severely mismatched with the actual heat transfer capacity of the liquid cooling system. This leads to overshoot oscillations in battery temperature or runaway temperature rise rate, which seriously affects the temperature uniformity and control accuracy of the battery pack under wide temperature range conditions. Summary of the Invention

[0003] This application aims to address at least one of the technical problems existing in the prior art.

[0004] Therefore, the purpose of this application is to propose a liquid cooling temperature control method for power batteries, which aims to solve the problem of liquid cooling control mismatch caused by sudden changes in the physical properties of the coolant.

[0005] This application proposes a liquid-cooled temperature control method for power batteries, including: Obtain the first operating parameters of the power battery and the second operating parameters of the liquid cooling system; The operating mode of the power battery is determined based on the first operating parameters, and the dynamic thermal management requirement index corresponding to the operating mode is obtained. Obtain the thermal coupling model of the power battery, which includes the heat exchange equations between the power battery and the coolant; The feedforward heat transfer coefficient between the power battery and the coolant is calculated based on the second operating parameter. Based on the heat exchange equation and according to the preset recursive least squares algorithm, the feedback heat transfer coefficient between the power battery and the coolant is determined. Based on the feedforward heat transfer coefficient and the feedback heat transfer coefficient, the target equivalent convective heat transfer coefficient of the power battery is generated, and the thermal coupling model is optimized according to the target equivalent convective heat transfer coefficient. Based on a preset multi-objective optimization algorithm and according to dynamic thermal management requirements, optimized thermal coupling model, and coolant inlet temperature and volumetric flow rate, the optimal control parameters of the liquid cooling system are determined. The liquid cooling system is controlled based on the optimal control parameters.

[0006] In some embodiments of this application, the first operating parameters include the average temperature, state of charge, and charge / discharge rate of the power battery. The process of determining the operating mode of the power battery based on the first operating parameters and obtaining the dynamic thermal management requirement indicators corresponding to the operating mode includes: defining a first fuzzy subset of the average temperature and determining a first membership function corresponding to the first fuzzy subset; defining a second fuzzy subset of the state of charge and determining a second membership function corresponding to the second fuzzy subset; defining a third fuzzy subset of the charge / discharge rate and determining a third membership function corresponding to the third fuzzy subset; determining the triggering rules and corresponding triggering intensities of the first operating parameters based on the first fuzzy subset, the first membership function, the second fuzzy subset, the second membership function, the third fuzzy subset, and the third membership function, according to a preset fuzzy rule base; calculating the operating mode index value using the centroid method based on the triggering rules and the corresponding triggering intensities, and determining the operating mode based on the operating mode index value; and determining the dynamic thermal management requirement indicators based on the operating mode according to a preset thermal management requirement mapping table.

[0007] In some embodiments of this application, generating the target equivalent convective heat transfer coefficient of the power battery based on the feedforward heat transfer coefficient and the feedback heat transfer coefficient includes: determining the flow stability factor of the coolant based on a second operating parameter; determining the first weighting coefficient of the feedforward heat transfer coefficient and the second weighting coefficient of the feedback heat transfer coefficient based on the flow stability factor; and calculating the target equivalent convective heat transfer coefficient based on the feedforward heat transfer coefficient, the first weighting coefficient, the feedback heat transfer coefficient, and the second weighting coefficient.

[0008] In some embodiments of this application, the sum of the first weighting coefficient and the second weighting coefficient is 1. Determining the first weighting coefficient of the feedforward heat transfer coefficient and the second weighting coefficient of the feedback heat transfer coefficient based on the flow stability factor includes: increasing the second weighting coefficient and decreasing the first weighting coefficient in response to the flow stability factor indicating that the flow state of the coolant tends to be stable; and decreasing the second weighting coefficient and increasing the first weighting coefficient in response to the flow stability factor indicating that the flow state of the coolant tends to be unstable.

[0009] In some embodiments of this application, the second operating parameters include the Reynolds number and critical Reynolds number of the coolant. The optimal control parameters of the liquid cooling system are determined based on a preset multi-objective optimization algorithm and according to dynamic thermal management requirements, the optimized thermal coupling model, the inlet temperature and volumetric flow rate of the coolant. This includes: pausing the optimization process and executing a boundary layer refresh scan procedure in response to a Reynolds number greater than the critical Reynolds number; and continuing iterative optimization after the boundary layer refresh scan procedure has been completed.

[0010] In some embodiments of this application, pausing optimization and executing a boundary layer refresh scan procedure includes: applying flow perturbation or temperature perturbation to the coolant to obtain temperature rise response data of the power battery; calculating and updating the target equivalent convective heat transfer coefficient based on the temperature rise response data; and further optimizing the thermal coupling model based on the updated target equivalent convective heat transfer coefficient.

[0011] In some embodiments of this application, after the boundary layer refresh scan procedure is completed, iterative optimization continues, including: recording the iterative optimization solution at the pause time; resetting the preset multi-objective optimization algorithm based on the iterative optimization solution; and using the further optimized thermal coupling model as the prediction model, continuing iterative optimization with the reset preset multi-objective optimization algorithm.

[0012] In some embodiments of this application, a flow rate disturbance or temperature disturbance is applied to the coolant to obtain temperature rise response data of the power battery. The target equivalent convective heat transfer coefficient is calculated and updated based on the temperature rise response data. This includes: presetting a disturbance sequence; applying temperature disturbance or flow rate disturbance to the coolant according to the disturbance sequence; collecting temperature rise response data of the power battery during the duration of each disturbance step and establishing a thermal balance equation before and after the disturbance; calculating the corresponding equivalent convective heat transfer coefficient disturbance value according to the thermal balance equation corresponding to each disturbance step, and calculating the average disturbance value of the equivalent convective heat transfer coefficient of the disturbance sequence to obtain the updated target equivalent convective heat transfer coefficient.

[0013] In some embodiments of this application, the optimal control parameters of the liquid cooling system are determined based on a preset multi-objective optimization algorithm and according to dynamic thermal management requirements, optimized thermal coupling model, inlet temperature and volumetric flow rate of coolant. This includes: obtaining the control parameter solution set at the end of the iteration; and determining the optimal control parameter solution in the control parameter solution set according to the approximation ideal solution method to obtain the optimal control parameters.

[0014] In some embodiments of this application, the liquid cooling system is controlled according to optimal control parameters, including: using a PID control algorithm to control the temperature of the coolant based on the optimal control parameters, and using a variable frequency closed-loop control algorithm to control the flow rate of the coolant.

[0015] The liquid-cooled temperature control method for power batteries disclosed in this application solves the technical problem in existing technologies where, when the coolant flow state undergoes a sudden change from laminar to turbulent, the heat transfer coefficient identified solely by historical data lags significantly behind the actual change in heat transfer capacity. It achieves real-time and accurate tracking of the equivalent convective heat transfer coefficient, providing a precise model basis for generating control strategies for the liquid cooling system. It also solves the technical problem in existing technologies where the control command model is based on a fixed heat transfer coefficient, leading to a mismatch between control commands and actual heat transfer capacity, and achieves dynamic adaptation of the control strategy to the actual physical state of the liquid cooling system. Furthermore, it addresses the technical problem that, when the flow state undergoes a drastic change, relying solely on observer weighted fusion may still result in a brief identification blind zone, causing the optimization algorithm to perform optimization based on transitional state data. Through active perturbation and real-time response verification, it ensures the accuracy of the heat transfer coefficient at the moment of the flow state change and the robustness of the optimization algorithm.

[0016] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0017] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a schematic flowchart of a liquid cooling temperature control method for a power battery according to some embodiments of this application; Figure 2 This is a flowchart illustrating a liquid-cooled temperature control method for a power battery according to other embodiments of this application; Figure 3 This is a flowchart illustrating a liquid-cooled temperature control method for a power battery according to other embodiments of this application; Figure 4 This is a flowchart illustrating a liquid-cooled temperature control method for a power battery according to other embodiments of this application; Figure 5 This is a flowchart illustrating a liquid-cooled temperature control method for a power battery according to other embodiments of this application. Detailed Implementation

[0018] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0019] This application proposes a liquid-cooled temperature control method for power batteries, as detailed below. Figure 1 This application describes a liquid-cooled temperature control method for a power battery according to embodiments of the present application. The liquid-cooled temperature control method for a power battery includes: S1. Obtain the first operating parameters of the power battery and the second operating parameters of the liquid cooling system.

[0020] The first operating parameter of the power battery includes the average temperature of the power battery. Maximum temperature Maximum temperature difference State of charge (SOC) and charge / discharge rate (C).

[0021] Specifically, the power battery includes a battery pack, a battery management system, a current and voltage detection unit, and a temperature sensor array installed inside the battery pack. The battery pack is composed of multiple individual cells, and the temperature sensor array can collect the surface temperature of each individual cell at a sampling frequency of not less than 10Hz. The current and voltage detection unit can collect the total voltage of the battery pack at a frequency synchronized with the temperature sampling. and total current The battery management system can obtain the state of charge (SOC) of the battery pack.

[0022] Furthermore, the charge / discharge rate C of the battery pack is calculated using the following formula: ;in, The rated capacity of the battery pack can be read from the battery parameter table stored in the battery management system.

[0023] Surface temperature of a single cell By performing a moving average filter, the average temperature of the battery pack can be calculated. Maximum temperature Minimum temperature and maximum temperature difference Specifically, the moving average filter uses a sliding window of length n (e.g., n=5) to perform an arithmetic average of n consecutive sampled values ​​of the surface temperature of each individual cell, thus obtaining the filtered individual cell temperature. Furthermore, Average temperature: ; Maximum temperature: ; Minimum temperature: ; Maximum temperature difference: ; Where N is the total number of individual cells.

[0024] Average temperature Maximum temperature Maximum temperature difference State of charge (SOC) and charge / discharge rate It constitutes a set of characteristic parameters that characterize the current thermal state of the battery pack.

[0025] The second operating parameter of the liquid cooling system includes the viscosity of the coolant. Coolant density Coolant flow rate Coolant inlet temperature and coolant flow rate .

[0026] Specifically, the liquid cooling system includes a viscosity sensor, a pressure sensor, a flow sensor, and a temperature sensor installed at the inlet of the liquid cooling system. The viscosity sensor can collect the viscosity of the coolant. The pressure sensor can collect the coolant pressure and calculate the coolant density from the coolant pressure. The flow sensor can collect the coolant flow rate. The temperature sensor can collect the coolant inlet temperature. .

[0027] Furthermore, the hydraulic diameter can be calculated based on the flow channel geometry parameters of the cooling system. and circulation cross-sectional area And further calculate the coolant flow rate. .

[0028] Furthermore, the Reynolds number Re of the coolant is: The Prandtl number Pr is: ;in, The specific heat capacity of the coolant. The thermal conductivity of the coolant can be determined by the coolant inlet temperature. The specific heat capacity and thermal conductivity of the coolant can be obtained from the coolant inlet temperature. The properties are obtained by consulting the property tables. These property tables characterize the key physical parameters of the coolant under different conditions and can be obtained through calibration experiments or based on theoretical data about the coolant.

[0029] S2. Determine the operating mode of the power battery based on the first operating parameters, and obtain the dynamic thermal management requirement index corresponding to the operating mode.

[0030] In some embodiments, step S2 is implemented using a fuzzy inference method. Specifically, a fuzzy inference rule base is pre-defined based on expert experience, and fuzzy inference and defuzzification operations are performed through a fuzzy logic controller. Taking the first operating parameter as input, the specific operating mode of the current power battery and the dynamic thermal management requirement indicators corresponding to the current operating mode are output. The dynamic thermal management requirement indicators include the target temperature, the maximum allowable temperature rise rate, and the maximum allowable temperature difference.

[0031] Among them, the operating modes of the power battery include at least the low temperature high rate charging condition, the low temperature low rate discharging condition, the high temperature high speed driving condition, the normal temperature regular charging and discharging condition, and the extreme environment start-up condition. These five operating conditions cover the typical scenarios with the greatest differences in the heat generation characteristics and safety risks of the power battery within a wide temperature range: low temperature high rate charging corresponds to the state with the highest risk of lithium plating, low temperature low rate discharging corresponds to the state with limited output capacity, high temperature high speed driving corresponds to the state of continuous high rate heat generation, normal temperature regular charging and discharging serves as the benchmark reference condition, and extreme environment start-up corresponds to the special transient process of recovery from a frozen state.

[0032] In some embodiments of this application, the first operating parameters include the average temperature, state of charge, and charge / discharge rate of the power battery. The process of determining the operating mode of the power battery based on the first operating parameters and obtaining the dynamic thermal management requirement indicators corresponding to the operating mode includes: defining a first fuzzy subset of the average temperature and determining a first membership function corresponding to the first fuzzy subset; defining a second fuzzy subset of the state of charge and determining a second membership function corresponding to the second fuzzy subset; defining a third fuzzy subset of the charge / discharge rate and determining a third membership function corresponding to the third fuzzy subset; determining the triggering rules and corresponding triggering intensities of the first operating parameters based on the first fuzzy subset, the first membership function, the second fuzzy subset, the second membership function, the third fuzzy subset, and the third membership function, according to a preset fuzzy rule base; calculating the operating mode index value using the centroid method based on the triggering rules and the corresponding triggering intensities, and determining the operating mode based on the operating mode index value; and determining the dynamic thermal management requirement indicators based on the operating mode according to a preset thermal management requirement mapping table.

[0033] Specifically, a fuzzy logic controller is used to implement the fuzzy inference method. The fuzzy logic controller employs a Mamdani-type fuzzy inference system, and the fuzzy inference method is as follows: Figure 2 As shown, the specific implementation process is as follows: Will SOC and charge / discharge rate As input variables for the fuzzy logic controller, multiple fuzzy subsets and their membership functions are defined for each input variable: Average temperature The first fuzzy subset is defined as {NB (low temperature), NS (medium temperature), ZO (high temperature)}, and the corresponding first membership function is a Gaussian function. Where x is the average temperature of the battery pack, and the parameter is... and Tuning based on experimental data; The second fuzzy subset of the State of Charge (SOC) is defined as {L_low, M_medium, H_high}, and the corresponding second membership function adopts a triangular function. The third fuzzy subset of the charge / discharge rate C is defined as {S small rate, M medium rate, B large rate}, and the corresponding third membership function adopts the trapezoidal function.

[0034] The default fuzzy rule base contains 27 rules, each in the form of: IF The output operating mode is D. Here, A, B, and C are fuzzy subsets of the input variables (i.e., average temperature, state of charge, and charge / discharge rate), and D is a fuzzy subset of the output operating mode.

[0035] Based on the aforementioned preset fuzzy rule base, and using the feature parameters obtained in step S1 (i.e., average temperature, state of charge, and charge / discharge rate), the triggering rules and corresponding triggering intensity are determined.

[0036] Defuzzification employs the centroid method, and the accurate operating condition mode index value is calculated using the following formula:

[0037] Where z is the operating mode triggered by the input variable; The output membership function represents the trigger strength (or weight) corresponding to each triggered operating condition mode; Mode is the defuzzified, precise output value. The operating condition mode index value is determined based on the defuzzified, precise output value through rounding or other methods.

[0038] The operating mode is determined based on the operating mode index value, and the preset thermal management demand mapping table is queried to determine the dynamic thermal management demand index, and the corresponding target temperature is output. Maximum permissible rate of temperature rise and maximum permissible temperature difference The thermal management requirement mapping table can be obtained by fitting battery life degradation experimental data.

[0039] In some embodiments, determining the operating mode of the power battery based on a first operating parameter includes: presetting a wide temperature range threshold for the battery pack, whereby the wide temperature range threshold includes a low-temperature freezing threshold, a low-temperature inefficiency threshold, a high-temperature warning threshold, and a high-temperature danger threshold; comparing the average temperature of the power battery with the aforementioned thresholds to determine the current temperature range to which the power battery belongs. In this embodiment, hard threshold discrimination can quickly pinpoint the macroscopic temperature range (such as freezing zone, inefficiency zone, high-temperature zone, etc.) of the power battery, providing a reduced initial domain of discourse or operating condition prediction for the fuzzy logic controller.

[0040] S3. Obtain the thermal coupling model of the power battery, which includes the heat exchange equations between the power battery and the coolant.

[0041] Specifically, the thermal coupling model of the power battery can be established based on the lumped mass method. The thermal coupling model includes the calculation formula for the heat generation rate q of a single cell: And the heat exchange equation between the power battery and the coolant: .in, t is the open-circuit voltage of the power battery; T is the real-time temperature of the power battery; q is the battery heat generation rate, in W, which represents the heat generated by a single battery cell per unit time. This is the equivalent convective heat transfer coefficient; m is the coolant temperature; m is the mass of the power battery. Where A is the specific heat capacity of the power battery, and A is the effective heat exchange area.

[0042] S4. The feedforward heat transfer coefficient between the power battery and the coolant is calculated based on the second operating parameters.

[0043] As described in step S1, the second operating parameter includes coolant viscosity. Coolant density Coolant flow rate Coolant inlet temperature Coolant flow rate The Reynolds number Re and Prandtl number Pr of the coolant.

[0044] Specifically, in some embodiments, such as Figure 3 As shown, the feedforward heat transfer coefficient Based on the pre-defined flow regime-Nusser number correlation The calculation is obtained. The Nusselt number correlation is determined based on the coolant flow pattern. like For laminar flow, the Nusselt number correlation is: ; like For turbulent flow, the Nusselt number correlation is: ; Feedforward heat transfer coefficient The calculation formula is: .

[0045] In other embodiments, in the feedforward calculation stage, a pre-experimentally calibrated three-dimensional lookup table model of coolant properties-flow regime-heat transfer coefficient can be used to replace the flow regime-Nusser number correlation, and the feedforward heat transfer coefficient can be obtained directly by querying the multidimensional interpolation table in real time.

[0046] S5. Based on the heat exchange equation and according to the preset recursive least squares algorithm, determine the feedback heat transfer coefficient between the power battery and the coolant.

[0047] Specifically, such as Figure 3 As shown, the feedback heat transfer coefficient The pre-defined recursive least squares method is used for online identification. The specific process of the pre-defined recursive least squares method is as follows: The heat exchange equation between the power battery and the coolant Discretize into difference form:

[0048] in, ; ; ; Where k is the index of the discrete sampling time, representing the time corresponding to the kth sampling period; The sampling period; The output quantity is a known quantity consisting of the current temperature change rate and the heat generation rate at the previous moment. Its physical meaning is the net heat power that needs to be dissipated through heat exchange at the current moment, which is the observable system output. The input quantity is a known quantity consisting of the difference between the average battery temperature and the coolant temperature at the previous moment, i.e., the temperature difference driving force of the current heat exchange process, which is used as the input regression vector. The parameter to be identified is the feedback heat transfer coefficient to be identified. The real-time estimate at time k.

[0049] The estimated values ​​of the parameters to be identified are updated using a pre-defined recursive least squares algorithm via a formula: Estimated values ​​of parameters to be identified:

[0050] Among them, gain , covariance , in, This is the forgetting factor, which can take a value of 0.98. The heat transfer coefficient is calculated using the above recursive method and output in real time. .

[0051] The pre-defined recursive least squares method serves as the online identification algorithm, running continuously without a defined termination condition. It outputs the latest estimated feedback heat transfer coefficient at each sampling time.

[0052] S6. Based on the feedforward heat transfer coefficient and the feedback heat transfer coefficient, generate the target equivalent convective heat transfer coefficient of the power battery, and optimize the thermal coupling model according to the target equivalent convective heat transfer coefficient.

[0053] Specifically, the feedback heat transfer coefficient With the feedforward heat transfer coefficient The input is fed into an adaptive weighted fusion observer, which then feeds back the heat transfer coefficient. With the feedforward heat transfer coefficient Perform weight allocation, generate and output the target equivalent convective heat transfer coefficient. Wherein, the target equivalent convective heat transfer coefficient is the corrected equivalent convective heat transfer coefficient. The adaptive weighted fusion observer is a flow-aware feedforward-feedback coupled heat transfer coefficient observer.

[0054] Using the target equivalent convective heat transfer coefficient Replace the equivalent convective heat transfer coefficient in the original heat exchange equation The heat exchange equations in the thermal coupling model of the power battery are updated to obtain the real-time updated thermal coupling model.

[0055] In some embodiments of this application, generating the target equivalent convective heat transfer coefficient of the power battery based on the feedforward heat transfer coefficient and the feedback heat transfer coefficient includes: determining the flow stability factor of the coolant based on a second operating parameter; determining the first weighting coefficient of the feedforward heat transfer coefficient and the second weighting coefficient of the feedback heat transfer coefficient based on the flow stability factor; and calculating the target equivalent convective heat transfer coefficient based on the feedforward heat transfer coefficient, the first weighting coefficient, the feedback heat transfer coefficient, and the second weighting coefficient.

[0056] Specifically, such as Figure 3 As shown, the feedback heat transfer coefficient With the feedforward heat transfer coefficient The input is fed into an adaptive weighted fusion observer, which adjusts the flow stability factor accordingly. Perform weight allocation, generate and output the target equivalent convective heat transfer coefficient. The first weighting coefficient is ( The second weighting coefficient is... .

[0057] Flow stability factor Based on the rate of change of the Reynolds number Re and its relationship with the critical Reynolds number The degree of proximity is dynamically adjusted. Specifically, the flow stability factor... The calculation and adaptive weighted fusion observer output process is as follows: Flow stability factor Defined as ,in and These are the weighting coefficients. The rate of change of Reynolds number, The absolute value of the distance between the current Reynolds number and the critical Reynolds number; when At that time, among them A preset threshold is used to determine if the flow pattern has crossed a critical region. The value approaches 0; when When it is large, The value decreases.

[0058] In some embodiments of this application, the sum of the first weighting coefficient and the second weighting coefficient is 1. Determining the first weighting coefficient of the feedforward heat transfer coefficient and the second weighting coefficient of the feedback heat transfer coefficient based on the flow stability factor includes: increasing the second weighting coefficient and decreasing the first weighting coefficient in response to the flow stability factor indicating that the flow state of the coolant tends to be stable; and decreasing the second weighting coefficient and increasing the first weighting coefficient in response to the flow stability factor indicating that the flow state of the coolant tends to be unstable.

[0059] When the rate of change of Reynolds number A larger value indicates that the coolant flow pattern is undergoing a rapid transition (e.g., a rapid transition from laminar to turbulent flow). In this case, the feedback heat transfer coefficient, recursively identified based on historical temperature data, is crucial. The inherent hysteresis prevents timely tracking of transient changes in actual heat transfer capacity, thus reducing its reliability. Therefore, the flow stability factor... The second weighting coefficient, which serves as the feedback component, decreases accordingly, affecting the output of the adaptive weighted fusion observer. More reliance on feedforward heat transfer coefficients calculated based on real-time physical property parameters The latter can reflect the theoretical heat transfer capacity under the current flow state in real time, thus ensuring that the corrected heat transfer coefficient still has high real-time accuracy when the flow state fluctuates drastically.

[0060] Furthermore, the feedback heat transfer coefficient of the adaptive weighted fusion observer is imported as the real-time output at the current moment, and weighted together with the synchronously calculated feedforward heat transfer coefficient. The two are updated and fused simultaneously in each control cycle. Therefore, the data flow in step S5 is updated cycle by cycle, and the feedback heat transfer coefficient can be continuously provided to the adaptive weighted fusion observer without the need for a termination condition.

[0061] S7. Based on the preset multi-objective optimization algorithm and according to the dynamic thermal management requirements, the optimized thermal coupling model, the inlet temperature and volumetric flow rate of the coolant, determine the optimal control parameters of the liquid cooling system.

[0062] Specifically, the optimal control parameters of the liquid cooling system are determined by using the dynamic thermal management demand index as the optimization objective, the updated thermal coupling model as the prediction model, and the inlet temperature and volumetric flow rate of the coolant as optimization variables, and by using a preset multi-objective optimization algorithm for iterative optimization.

[0063] Among them, such as Figure 4 As shown, the specific parameter settings and iterative optimization process of the preset multi-objective genetic algorithm NSGA-II are as follows: The population size is set to 100, and the optimization variable is the target coolant inlet temperature. and target coolant volume flow rate , The search scope is , The search scope is The search scope is determined by the hardware capabilities of the liquid cooling system.

[0064] The fitness function of a multi-objective genetic algorithm includes the following three objective functions: , is the integral of the temperature tracking deviation; The integral for the penalty of exceeding the temperature rise rate limit; The penalty points are for exceeding the temperature difference limit; in, This should be understood as: at some future moment within the prediction time domain. The maximum temperature difference of the battery pack predicted by the thermal coupling model optimized and updated in step S6. The maximum allowable temperature difference threshold output in step S2 The difference between the predicted and actual temperature differences. If the difference is less than or equal to zero, it means the predicted temperature difference is within the acceptable range, and no penalty is imposed. If the difference is greater than zero, it means the predicted temperature difference has exceeded the allowable range, and the integral of the excess portion over time constitutes the penalty for exceeding the temperature difference limit. This is used to drive optimization algorithms in multi-objective optimization to prioritize candidate solutions with better temperature difference control capabilities.

[0065] in, To predict the length of the time domain, The temperature at future moments is predicted by the thermally coupled model updated in step S6. This fitness function will predict the temperature sequence. With target temperature The cumulative deviation is used as one of the optimization objectives to drive a multi-objective genetic algorithm to find a way to make the future battery temperature as close as possible to the target temperature. Optimal control parameters .

[0066] In some embodiments, the multi-objective genetic algorithm employs simulated binary crossover and polynomial mutation operators, with the crossover probability set to 0.9 and the mutation probability set to 0.1. The iteration termination condition is reaching the maximum generation of 200 or the Pareto front change being less than a preset threshold for 20 consecutive generations. The preset threshold is determined through experimental calibration or engineering experience, and is generally 1%-5%.

[0067] In some embodiments of this application, the second operating parameters include the Reynolds number and critical Reynolds number of the coolant. The optimal control parameters of the liquid cooling system are determined based on a preset multi-objective optimization algorithm and according to dynamic thermal management requirements, the optimized thermal coupling model, the inlet temperature and volumetric flow rate of the coolant. This includes: pausing the optimization process and executing a boundary layer refresh scan procedure in response to a Reynolds number greater than the critical Reynolds number; and continuing iterative optimization after the boundary layer refresh scan procedure has been completed.

[0068] Specifically, such as Figure 4 As shown, during the iterative optimization process using a multi-objective genetic algorithm, when the Reynolds number Re is detected to cross the critical Reynolds number... When the Reynolds number crosses a critical value, it signifies a sudden change in the flow field state. If optimization continues based on old data at this critical moment, the results will deviate significantly from reality, leading to optimization failure. Calling the boundary layer refresh scan program, by pausing optimization and refreshing the boundary layer data, can improve prediction accuracy.

[0069] In some embodiments of this application, pausing optimization and executing a boundary layer refresh scan procedure includes: applying flow perturbation or temperature perturbation to the coolant to obtain temperature rise response data of the power battery; calculating and updating the target equivalent convective heat transfer coefficient based on the temperature rise response data; and further optimizing the thermal coupling model based on the updated target equivalent convective heat transfer coefficient.

[0070] Specifically, the liquid-cooled pump is forced to perform a preset small-amplitude stepped flow disturbance, and the corrected equivalent convective heat transfer coefficient is back-verified and updated based on the real-time temperature rise response data. And based on the updated The initial values ​​of individuals in the multi-objective genetic algorithm population are reset, and iterative optimization continues until the termination condition is met. The solution with the highest overall satisfaction is selected from the Pareto front solution set as the optimal control parameters at the current moment. .

[0071] Specifically, the real-time temperature rise response data refers to the sequence of temperature changes of each individual cell over time, collected by the temperature sensor array during each perturbation step, during the execution of the boundary layer refresh scan program. This directly reflects the recorded temperature rise rate of each individual cell. The real-time temperature rise response data can be used to subsequently substitute into the heat exchange equation to solve for the equivalent convective heat transfer coefficient corresponding to each perturbation step.

[0072] In some embodiments of this application, a flow rate disturbance or temperature disturbance is applied to the coolant to obtain temperature rise response data of the power battery. The target equivalent convective heat transfer coefficient is calculated and updated based on the temperature rise response data. This includes: presetting a disturbance sequence; applying temperature disturbance or flow rate disturbance to the coolant according to the disturbance sequence; collecting temperature rise response data of the power battery during the duration of each disturbance step and establishing a thermal balance equation before and after the disturbance; calculating the corresponding equivalent convective heat transfer coefficient disturbance value according to the thermal balance equation corresponding to each disturbance step, and calculating the average disturbance value of the equivalent convective heat transfer coefficient of the disturbance sequence to obtain the updated target equivalent convective heat transfer coefficient.

[0073] Specifically, the liquid cooling pump is forced to execute a preset small-amplitude stepped flow rate disturbance sequence, the disturbance sequence being as follows: ,in For the preset traffic increment, The number of perturbation steps, The actual coolant volume flow rate of the liquid cooling system at the current moment when the boundary layer refresh scan program is started.

[0074] During each perturbation step, battery temperature response data is acquired via a temperature sensor array, and the temperature rise rate of each individual cell is recorded. According to the heat exchange equations before and after the disturbance Inverse calculation of the equivalent convective heat transfer coefficient for each disturbance step ;right The perturbation steps are calculated as follows The updated target equivalent convective heat transfer coefficient is obtained by performing an arithmetic average. .

[0075] In some embodiments of this application, after the boundary layer refresh scan procedure is completed, iterative optimization continues, including: recording the iterative optimization solution at the pause time; resetting the preset multi-objective optimization algorithm based on the iterative optimization solution; and using the further optimized thermal coupling model as the prediction model, continuing iterative optimization with the reset preset multi-objective optimization algorithm.

[0076] Specifically, using the updated target equivalent convective heat transfer coefficient Replace the target equivalent convective heat transfer coefficient output in step S6 Then, based on the iterative optimization solution near the pause time, new initial values ​​of individuals in the population are randomly generated, the iteration counter of the preset multi-objective genetic algorithm is reset, and then the iterative optimization process of step S7 is continued.

[0077] like Figure 4As shown, in some embodiments of this application, the optimal control parameters of the liquid cooling system are determined based on a preset multi-objective optimization algorithm and according to dynamic thermal management requirements, optimized thermal coupling model, inlet temperature and volumetric flow rate of coolant. This includes: obtaining the control parameter solution set at the end of the iteration; and determining the optimal control parameter solution in the control parameter solution set according to the approximation ideal solution method to obtain the optimal control parameters.

[0078] Specifically, the optimal control parameters are selected from the control parameter solution set at the end of the iteration (i.e., the Pareto front solution set). The specific process is as follows: After the multi-objective genetic algorithm terminates its iterations, a set of Pareto optimal solutions is obtained. The Top-Optimal Solution Ranking Method (TOPSIS) is used for multi-attribute decision-making, and a decision matrix is ​​constructed. Each row corresponds to a Pareto solution, and each column corresponds to three objective functions. , , The value of the decision matrix is ​​then normalized to obtain the normalized matrix. Determine the ideal solution and negative ideal solution Calculate the Euclidean distance between each solution and the positive ideal solution. The Euclidean distance between the sum and the negative ideal solution ; Calculate relative proximity Choose the solution with the highest relative similarity. As the optimal control parameter at the current moment.

[0079] S8. Control the liquid cooling system according to the optimal control parameters.

[0080] Specifically, the optimal control parameters generated in step S7 The signal is sent to the controller of the liquid cooling system, which adjusts the opening of the electronic expansion valve or the power of the refrigeration unit to control the coolant temperature. At the same time, adjust the speed of the electronic water pump to control the coolant flow rate. This enables adaptive and precise temperature control of the power battery pack under wide temperature range conditions.

[0081] In some embodiments of this application, the liquid cooling system is controlled according to optimal control parameters, including: using a PID control algorithm to control the temperature of the coolant based on the optimal control parameters, and using a variable frequency closed-loop control algorithm to control the flow rate of the coolant.

[0082] Specifically, such as Figure 5 As shown, a PID controller is used to control the coolant temperature. The specific process is as follows: The control law of the PID controller is , in, This represents the temperature deviation. To measure the coolant temperature, , , For PID parameters, the controller output Adjust the opening of the expansion valve of the refrigeration unit or the power of the heater.

[0083] The coolant flow rate is controlled using a variable frequency closed-loop control algorithm, and the specific process is as follows: The formula for calculating the pump drive frequency is as follows: , in, For the water pump drive frequency, As the reference frequency, This is the proportionality coefficient. To measure the actual flow rate, the coolant flow rate was adjusted by changing the water pump speed. The controller performs the above adjustments in a cycle of no less than 1 Hz until it receives the updated optimal control parameters for the next moment.

[0084] It should be noted that the lowest temperature obtained in step S1 The application scenarios are as follows: when the power battery is in an extremely cold start-up condition, the lowest temperature Used with freezing temperature Compare and determine if there is a risk of localized freezing inside the power battery. If Then regardless of average temperature Regardless of the value, the low temperature protection mode will be forcibly triggered. This logic is implicit in the "extreme environment start-up condition" identification rule in step S2.

[0085] Maximum temperature difference The application scenarios are as follows: maximum temperature difference The input is directly fed into the multi-objective iterative optimization step S7 as the basis for real-time calculation of the temperature difference exceedance penalty integral: ,in The predictions were obtained from the thermally coupled model optimized and updated in step S6, and This is obtained from the operating condition identification process in step S2. Maximum temperature difference. The temperature control strategy should focus on both the average temperature and the internal temperature uniformity constraints of the power battery.

[0086] In the liquid cooling temperature control method for power batteries proposed in this application, the coupled heat transfer coefficient is observed based on flow regime sensing. Building upon the traditional recursive least squares online identification of the feedback heat transfer coefficient, coolant property parameters acquired in real-time by viscosity and pressure sensors are introduced. The Reynolds number is calculated in real-time, and it is determined whether the flow regime crosses a critical value. An adaptive weighted fusion is performed on the feedback heat transfer coefficient identified by recursive least squares and the feedforward heat transfer coefficient calculated based on the flow regime-Nusser number correlation, according to the flow regime stability factor. An adaptive adjustment strategy is generated based on a multi-objective genetic algorithm and fused with flow regime sensing. The target temperature, allowable temperature rise rate, and maximum allowable temperature difference output by the operating condition identification module are used as optimization objectives. The battery thermal coupling model updated by the flow regime sensing observer is used as the prediction model. The optimal coolant inlet temperature and flow rate are determined through iterative optimization using a multi-objective genetic algorithm. An online correction mechanism is established by boundary layer refresh scanning. During the iterative optimization process of multi-objective genetic algorithm, the change of Reynolds number is monitored in real time. When the Reynolds number is detected to cross the critical Reynolds number, the current optimization process is immediately paused and the liquid cooling pump is forced to execute a preset small step flow disturbance sequence. The equivalent convective heat transfer coefficient is back-verified and updated based on the real-time temperature rise response data. At the same time, the initial value of the population is reset and the optimization continues.

[0087] The liquid-cooled temperature control method for power batteries disclosed in this application solves the technical problem in existing technologies where, when the coolant flow state undergoes a sudden change from laminar to turbulent, the heat transfer coefficient identified solely by historical data lags significantly behind the actual change in heat transfer capacity. It achieves real-time and accurate tracking of the equivalent convective heat transfer coefficient, providing a precise model basis for generating control strategies for the liquid cooling system. It also solves the technical problem in existing technologies where the control command model is based on a fixed heat transfer coefficient, leading to a mismatch between control commands and actual heat transfer capacity, and achieves dynamic adaptation of the control strategy to the actual physical state of the liquid cooling system. Furthermore, it addresses the technical problem that, when the flow state undergoes a drastic change, relying solely on observer weighted fusion may still result in a brief identification blind zone, causing the optimization algorithm to perform optimization based on transitional state data. Through active perturbation and real-time response verification, it ensures the accuracy of the heat transfer coefficient at the moment of the flow state change and the robustness of the optimization algorithm.

[0088] In the description of this application, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0089] In the description of this application, "first feature" and "second feature" may include one or more of the features.

[0090] In the description of this application, "multiple" means two or more.

[0091] In the description of this application, the first feature being "above" or "below" the second feature may include the first and second features being in direct contact, or the first and second features being in contact through another feature between them.

[0092] In the description of this application, the terms "above," "over," and "on top" for the first feature and the second feature include the first feature being directly above or diagonally above the second feature, or simply indicate that the first feature is at a higher horizontal level than the second feature.

[0093] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0094] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

Claims

1. A liquid-cooled temperature control method for a power battery, characterized in that, include: Obtain the first operating parameters of the power battery and the second operating parameters of the liquid cooling system; The operating mode of the power battery is determined based on the first operating parameters, and the dynamic thermal management requirement index corresponding to the operating mode is obtained. Obtain the thermal coupling model of the power battery, wherein the thermal coupling model includes the heat exchange equations between the power battery and the coolant; The feedforward heat transfer coefficient between the power battery and the coolant is calculated based on the second operating parameters. Based on the heat exchange equation and according to the preset recursive least squares algorithm, the feedback heat transfer coefficient between the power battery and the coolant is determined; Based on the feedforward heat transfer coefficient and the feedback heat transfer coefficient, the target equivalent convective heat transfer coefficient of the power battery is generated, and the thermal coupling model is optimized according to the target equivalent convective heat transfer coefficient. Based on a preset multi-objective optimization algorithm and according to the dynamic thermal management requirement index, the optimized thermal coupling model, the inlet temperature and volumetric flow rate of the coolant, the optimal control parameters of the liquid cooling system are determined. The liquid cooling system is controlled according to the optimal control parameters.

2. The liquid cooling temperature control method for a power battery according to claim 1, characterized in that, The first operating parameters include the average temperature, state of charge, and charge / discharge rate of the power battery. The step of determining the operating mode of the power battery based on the first operating parameters and obtaining the dynamic thermal management requirement indicators corresponding to the operating mode includes: Define a first fuzzy subset of the average temperature and determine the first membership function corresponding to the first fuzzy subset; Define a second fuzzy subset of the charged states, and determine the second membership function corresponding to the second fuzzy subset; Define a third fuzzy subset of the charge / discharge rate and determine the third membership function corresponding to the third fuzzy subset; Based on the first fuzzy subset, the first membership function, the second fuzzy subset, the second membership function, the third fuzzy subset, and the third membership function, the triggering rules and corresponding triggering strengths of the first running parameters are determined according to a preset fuzzy rule library. Based on the triggering rules and the corresponding triggering intensity, the working condition mode index value is calculated using the centroid method, and the working condition mode is determined according to the working condition mode index value. Based on a preset thermal management demand mapping table, the dynamic thermal management demand indicators are determined according to the operating conditions.

3. The liquid cooling temperature control method for a power battery according to claim 1, characterized in that, The step of generating the target equivalent convective heat transfer coefficient of the power battery based on the feedforward heat transfer coefficient and the feedback heat transfer coefficient includes: Based on the second operating parameters, the flow stability factor of the coolant is determined; Based on the flow stability factor, determine the first weighting coefficient of the feedforward heat transfer coefficient and the second weighting coefficient of the feedback heat transfer coefficient; The target equivalent convective heat transfer coefficient is calculated based on the feedforward heat transfer coefficient, the first weighting coefficient, the feedback heat transfer coefficient, and the second weighting coefficient.

4. The liquid cooling temperature control method for a power battery according to claim 3, characterized in that, The sum of the first weighting coefficient and the second weighting coefficient is 1, wherein determining the first weighting coefficient of the feedforward heat transfer coefficient and the second weighting coefficient of the feedback heat transfer coefficient based on the flow stability factor includes: In response to the flow stability factor indicating that the flow state of the coolant tends to be stable, the second weighting coefficient is increased and the first weighting coefficient is decreased; In response to the flow stability factor indicating that the flow state of the coolant tends to be unstable, the second weighting coefficient is decreased and the first weighting coefficient is increased.

5. The liquid cooling temperature control method for a power battery according to claim 1, characterized in that, The second operating parameters include the Reynolds number and critical Reynolds number of the coolant. The determination of the optimal control parameters for the liquid cooling system based on a preset multi-objective optimization algorithm and according to the dynamic thermal management requirements, the optimized thermal coupling model, and the inlet temperature and volumetric flow rate of the coolant includes: In response to the Reynolds number being greater than the critical Reynolds number, the optimization process is paused and a boundary layer refresh scan procedure is executed; after the boundary layer refresh scan procedure is completed, iterative optimization continues.

6. The liquid cooling temperature control method for a power battery according to claim 5, characterized in that, The pause optimization and execution of the boundary layer refresh scan procedure includes: Apply flow disturbance or temperature disturbance to the coolant to obtain temperature rise response data of the power battery; The target equivalent convective heat transfer coefficient is calculated and updated based on the temperature rise response data, and the thermal coupling model is further optimized based on the updated target equivalent convective heat transfer coefficient.

7. The liquid cooling temperature control method for a power battery according to claim 6, characterized in that, After the boundary layer refresh scan procedure is completed, the iterative optimization continues, including: Record the iterative optimization solution at each pause point; Based on the iterative optimization solution, the preset multi-objective optimization algorithm is reset; Using the further optimized thermal coupling model as the prediction model, the preset multi-objective optimization algorithm, after being reset, is used to continue iterative optimization.

8. The liquid cooling temperature control method for a power battery according to claim 6, characterized in that, The process of applying flow or temperature disturbances to the coolant, acquiring temperature rise response data of the power battery, and calculating and updating the target equivalent convective heat transfer coefficient based on the temperature rise response data includes: A preset perturbation sequence is used to apply temperature or flow perturbation to the coolant according to the perturbation sequence; During each disturbance step, the temperature rise response data of the power battery is collected, and the thermal balance equation before and after the disturbance is established. Based on the heat balance equation corresponding to each disturbance step, the corresponding equivalent convective heat transfer coefficient disturbance value is calculated, and the mean value of the equivalent convective heat transfer coefficient disturbance of the disturbance sequence is calculated to obtain the updated target equivalent convective heat transfer coefficient.

9. The liquid cooling temperature control method for a power battery according to claim 1, characterized in that, The determination of the optimal control parameters for the liquid cooling system based on a preset multi-objective optimization algorithm and according to the dynamic thermal management requirements, the optimized thermal coupling model, and the inlet temperature and volumetric flow rate of the coolant includes: Obtain the control parameter solution set at the time of iteration termination; The optimal control parameter solution in the control parameter solution set is determined by the approximation of the ideal solution method, so as to obtain the optimal control parameters.

10. The liquid cooling temperature control method for a power battery according to claim 1, characterized in that, The step of controlling the liquid cooling system according to the optimal control parameters includes: Based on the optimal control parameters, a PID control algorithm is used to control the temperature of the coolant, and a variable frequency closed-loop control algorithm is used to control the flow rate of the coolant.

Images