A battery thermal management system optimization method

Abstract

The present application relates to the technical field of optimization control of battery thermal management system, in particular to a kind of battery thermal management system optimization method.The method comprises the following steps: S1, obtaining the operating state parameters and battery aging state parameters of battery pack;Based on operating state parameters, construct a thermal characteristic prediction model to generate predicted maximum temperature and predicted maximum temperature difference;S2, set energy consumption optimization target and initial safety constraint;Combined with predicted maximum temperature and predicted maximum temperature difference, construct the Lagrange optimization function considering initial safety constraint;S3, based on battery aging state parameters, calculate battery aging factor;According to battery aging factor, dynamically correct initial safety constraint to generate corrected dynamic safety limit;The Lagrange optimization function is updated using the corrected dynamic safety limit, and the updated Lagrange optimization function is solved to output the optimal cooling system power instruction.The method realizes the dynamic self-adaptive adjustment of thermal management safety boundary.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of optimization control of battery thermal management system, in particular to a battery thermal management system optimization method. BACKGROUND

[0002] The battery thermal management system is a key system for maintaining the battery in a suitable temperature range, ensuring its performance, safety and life; its basic principle is to accurately control the heat production and heat dissipation balance of the battery pack through the cooling system such as water pump and fan, so as to control the maximum temperature and internal temperature difference of the battery within the designed range;

[0003] In order to ensure the safe and efficient operation of the battery throughout its life cycle, the thermal management system needs to obtain the running state parameters of the battery in real time, such as current, temperature, state of charge, and aging state parameters reflecting its long-term health level, such as state of health (SOH) and equivalent cycle number; however, as an electrochemical system, the internal heat production characteristics and its resistance to high temperature will change with use and aging; the existing thermal management technology mostly uses fixed safety threshold for control, that is, the upper limit of the maximum temperature and maximum temperature difference of the battery throughout its life cycle is constant; this static control strategy does not fully consider the deterioration of battery aging characteristics, when the battery is in the aging state, the fixed safety boundary may no longer be applicable, resulting in ineffective delay of battery life decay, and even potential safety hazards under extreme working conditions.

[0004] The above information disclosed in the above background section is only used to strengthen the understanding of the background of the present disclosure, therefore it can include information which does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0005] To solve the above technical problems, the present application discloses a battery thermal management system optimization method, in particular, the technical scheme of the present application is:

[0006] A battery thermal management system optimization method, comprising:

[0007] S1, obtaining the running state parameters and the battery aging state parameters of the battery pack; based on the running state parameters, constructing a thermal characteristic prediction model to generate a predicted maximum temperature and a predicted maximum temperature difference;

[0008] S2, setting an energy consumption optimization target and an initial safety constraint; combining the predicted maximum temperature and the predicted maximum temperature difference, constructing a Lagrange optimization function considering the initial safety constraint;

[0009] S3, calculating a battery aging factor based on the battery aging state parameter; dynamically revising the initial safety constraint according to the battery aging factor to generate a revised dynamic safety limit; updating the Lagrange optimization function by using the revised dynamic safety limit, and solving the updated Lagrange optimization function to output an optimal cooling system power instruction.

[0010] Preferably, S1 comprises:

[0011] Real-time acquisition of the charging and discharging current and the cell temperature of the battery pack; and real-time acquisition of the external environment temperature and the cooling liquid inlet temperature;

[0012] Based on the acquired charging and discharging current and cell temperature, the battery heat generation rate is calculated by using the corrected Bernard heat model.

[0013] Based on the acquired cell temperature, cooling liquid inlet temperature and external environment temperature, the battery heat dissipation rate is calculated by using the heat dissipation model determined through system identification experiment.

[0014] Preferably, S1 further comprises:

[0015] Based on the calculated battery heat generation rate and battery heat dissipation rate, a dynamic evolution prediction model of the battery temperature state is established.

[0016] The dynamic evolution prediction model generates a predicted maximum temperature by using energy conservation derivation, and generates a predicted maximum temperature difference by using a simplified linear state space model.

[0017] Preferably, S2 comprises:

[0018] The energy consumption optimization goal is set as minimizing the sum of the power of the cooling system water pump and the fan;

[0019] The initial safety constraint is set as that the maximum temperature of the battery pack does not exceed a preset maximum temperature safety upper limit, and the maximum temperature difference of the battery pack does not exceed a preset maximum temperature difference consistency upper limit.

[0020] Preferably, constructing the Lagrange optimization function comprises:

[0021] Based on the predicted maximum temperature and the maximum temperature safety upper limit, a temperature constraint function is constructed;

[0022] Based on the predicted maximum temperature difference and the maximum temperature difference consistency upper limit, a temperature difference constraint function is constructed;

[0023] The Lagrange optimization function is constructed in combination with the energy consumption optimization goal, the temperature constraint function and the temperature difference constraint function.

[0024] Preferably, the battery aging state parameter comprises a battery health state and an equivalent cycle number estimated by a battery management system.

[0025] The battery aging factor is calculated, comprising:

[0026] The battery aging factor is calculated based on the battery state of health and the equivalent cycle number through a preset aging factor model.

[0027] Preferably, the generation of the modified dynamic safety limit value comprises:

[0028] Based on the calculated battery aging factor, the highest temperature safety upper limit preset in the initial safety constraint is modified in a proportional adjustment manner to generate a modified highest temperature limit value.

[0029] Based on the calculated battery aging factor, the maximum temperature difference consistency upper limit preset in the initial safety constraint is modified in a proportional adjustment manner to generate a modified maximum temperature difference limit value.

[0030] Preferably, the solving comprises:

[0031] In each control period, the Karush-Kuhn-Tucker optimality condition corresponding to the updated Lagrange optimization function is solved to obtain an optimal power distribution.

[0032] The optimal power distribution is output to the water pump and fan driver of the cooling system as an optimal cooling system power instruction.

[0033] Compared with the prior art, the present application has the following beneficial effects:

[0034] 1. The method realizes dynamic adaptive adjustment of the thermal management safety boundary. The method can calculate an aging factor in real time according to the aging parameters such as the battery state of health and the equivalent cycle number estimated by the battery management system, and dynamically modify the safety limit values of the highest temperature and the maximum temperature difference accordingly. Compared with the traditional method of using fixed threshold values, this strategy of prospectively tightening the operating boundary can effectively delay the life attenuation of the aging battery while ensuring its safety.

[0035] 2. The method realizes whole life cycle optimization of the cooling system energy consumption. By constructing an optimization function with the minimization of the total power of the cooling water pump and fan as the target, and solving the optimality condition in each control period, the cooling system is ensured to always operate at the lowest power consumption under the premise of meeting the dynamic safety constraint. This enables the system to continuously and dynamically find the lowest energy consumption point, thereby achieving significant energy-saving effect in the whole life cycle of the battery.

[0036] 3、The method improves the control accuracy and response speed of the thermal management system. The method constructs a thermal characteristic prediction model based on real-time operating parameters, which can accurately estimate the battery maximum temperature and maximum temperature difference in a specific control period in the future. This prediction ability changes the control strategy from the traditional responsive to the forward-looking predictive control, avoiding control lag and improving the accuracy and robustness of the system in response to complex working condition changes.

[0037] 4、The method realizes the coordination and optimal balance among multiple targets. The method places the real-time working condition, long-term health status, system energy consumption and operation safety of the battery in a unified Lagrange optimization framework for solving. Through this systematic mathematical modeling method, the complex boundary constraint problem is transformed into an unconstrained function extremum problem, realizing the coordinated optimization and dynamic balance of the two core indicators of operation safety and system energy consumption. BRIEF DESCRIPTION OF DRAWINGS

[0038] The application will be further explained in conjunction with the accompanying drawings and examples:

[0039] Figure 1 is a flow chart of the method of the application. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical scheme and advantages of the application clearer, the application will be further described in detail below in conjunction with specific examples.

[0041] Example 1:

[0042] Please refer to Figure 1 A battery thermal management system optimization method, comprising:

[0043] S1, obtaining the operating state parameters and the battery aging state parameters of the battery pack; based on the operating state parameters, constructing a thermal characteristic prediction model to generate a predicted maximum temperature and a predicted maximum temperature difference;

[0044] S2, setting an energy consumption optimization target and an initial safety constraint; combining the predicted maximum temperature and the predicted maximum temperature difference, constructing a Lagrange optimization function considering the initial safety constraint;

[0045] S3, calculating a battery aging factor based on the battery aging state parameters; dynamically modifying the initial safety constraint according to the battery aging factor to generate a modified dynamic safety limit; updating the Lagrange optimization function using the modified dynamic safety limit, and solving the updated Lagrange optimization function to output an optimal cooling system power instruction;

[0046] The embodiment of the application provides a battery thermal management system optimization method, aiming at realizing dynamic optimal balance of battery operation safety and system energy consumption in the whole life cycle of the battery; the method constitutes a complete and self-consistent technical closed loop, and specific steps include:

[0047] S1, acquiring operation state parameters and battery aging state parameters of a battery pack, and constructing a thermal characteristic prediction model; this step is the basis of the whole optimization control strategy; the operation state parameters refer to real-time physical quantities reflecting current working conditions of the battery, such as current, state of charge, temperature and the like, and the operation state parameters provide instant input for the prediction model; the battery aging state parameters refer to quantitative indexes representing long-term health levels of the battery, such as state of health SOH and equivalent cycle number, and the battery aging state parameters provide a basis for subsequent dynamic adjustment of a safety boundary; based on the real-time collected operation state parameters, the embodiment constructs a thermal characteristic prediction model, and the core purpose of the thermal characteristic prediction model is to establish an accurate mathematical relationship between a control variable, that is, cooling system power, and future temperature states of the battery, that is, predicted maximum temperature and predicted maximum temperature difference; the differentiable characteristic of the model is a prerequisite for subsequent solving by using a gradient-based optimization algorithm;

[0048] S2, setting an energy consumption optimization target and an initial safety constraint, and constructing a Lagrange optimization function considering the initial safety constraint; this step aims to formalize actual engineering requirements into a standard mathematical optimization problem; the energy consumption optimization target refers to minimizing total power of cooling system actuators such as a water pump and a fan, and this is a core demand of system energy saving; the initial safety constraint refers to upper limits of the maximum temperature and the maximum temperature difference that must be complied with in the initial stage of the whole life of the battery according to battery design specifications and safety standards; for systematic solving of this optimization problem with inequality constraints, the embodiment constructs a Lagrange optimization function; the application logic of the function is that the function converts a complex boundary constraint problem into a function extreme value problem without constraints by introducing a Lagrange multiplier, greatly simplifies the solving process, and unifies the energy consumption target and the safety constraint in a single mathematical framework;

[0049] S3, calculating a battery aging factor based on the battery aging state parameter, dynamically revising the initial safety constraint, and solving the updated Lagrange optimization function to output an optimal cooling system power instruction; this step embodies the core adaptive and forward-looking capability of the application; first, based on the obtained battery aging state parameter, a comprehensive battery aging factor is calculated, which aims to quantify the evolution process of the battery from health to decline; second, the method uses the aging factor to dynamically revise the initial safety constraint set in step S2 to generate a revised dynamic safety limit; the technical motivation of this design is that as the battery ages, its heat generation characteristics and resistance to high temperature will deteriorate, and the fixed safety boundary is no longer applicable; by actively tightening the operating boundary, the battery life can be effectively delayed under the premise of ensuring safety; finally, the revised limit is used to update the Lagrange optimization function, and the function is solved to output the optimal cooling system power instruction, which is the optimal solution under the current working condition and the current battery health state. The instruction is sent to the cooling system actuator to complete a closed-loop control;

[0050] The embodiment establishes a closed-loop control method integrating prediction-optimization-adaptation by constructing a prediction model, setting an optimization target, and dynamically revising the safety constraint under the driving of the aging state; compared with the traditional heat management strategy using fixed threshold or responsive control, the application can dynamically and continuously find the lowest point of system energy consumption according to the real-time working condition and long-term health status of the battery under the premise of ensuring the absolute safety of the battery, thereby achieving significant energy-saving effect and effectively prolonging the service life of the battery throughout the life cycle of the battery.

[0051] Embodiment 2:

[0052] S1 includes:

[0053] The charging and discharging currents of the battery pack and the cell temperature are collected in real time; and the external environment temperature and the cooling liquid inlet temperature are collected in real time;

[0054] Based on the collected charging and discharging currents and cell temperature, the battery heat generation rate is calculated using the modified Bernard heat model;

[0055] Based on the collected cell temperature, cooling liquid inlet temperature and external environment temperature, the battery heat dissipation rate is calculated through the heat dissipation model determined by system identification experiment;

[0056] S1 further includes:

[0057] Based on the calculated battery heat generation rate and battery heat dissipation rate, a dynamic evolution prediction model of the battery temperature state is established;

[0058] The dynamic evolution prediction model generates a predicted maximum temperature using energy conservation and generates a predicted maximum temperature difference using a simplified linear state space model;

[0059] This embodiment is based on the embodiment 1, the heat characteristic prediction model construction process in step S1 is specified and optimized, so as to establish a high-precision prediction model which can be used for online optimization;

[0060] This embodiment collects the charging and discharging current and the cell temperature of the battery pack, and the external environment temperature and the cooling liquid inlet temperature in real time through the sensor; these are the basic physical quantities of the model input; among them, the charging and discharging current and the cell temperature are the core basis for calculating heat generation; and the external environment temperature and the cooling liquid inlet temperature are the key boundary conditions for calculating heat dissipation;

[0061] Based on the collected charging and discharging current and cell temperature, this embodiment uses the modified Bernard heat model to calculate the battery heat generation rate; the purpose of this step is to accurately quantify the intensity of the internal heat source of the battery; the mathematical expression of the model is:

[0062]

[0063] Among them, : total heat generation rate, unit: watt , which is calculated by the formula;

[0064] : real-time current, unit: ampere , which is the real-time collection value of the vehicle-mounted current sensor;

[0065] : battery equivalent internal resistance, unit: ohm , which is a function of the state of charge and the temperature , through the thermal characteristic experiment of a specific cell, such as hybrid pulse power characteristic test, HPPC, and using data fitting method for calibration, finally in the form of lookup table or fitting function stored in the controller for real-time calling;

[0066] : battery temperature, which is the real-time collection value of the internal temperature sensor of the battery pack; when calculating the equivalent internal resistance , Celsius or Kelvin can be used; when calculating the entropy heat term , in order to ensure physical accuracy, Kelvin must be used as the unit;

[0067] : entropy coefficient, unit: volt per Kelvin Similar to the equivalent resistance, it is obtained through experimental calibration and varies with temperature;

[0068] Based on the collected cell temperature, cooling liquid inlet temperature and external environment temperature, the battery heat dissipation rate is calculated through the heat dissipation model determined by system identification experiment; this step aims to establish a direct quantitative relationship between the cooling system execution power and the actual heat dissipation effect; its mathematical expression is:

[0069]

[0070] Among them, : total heat dissipation rate, unit: watt , calculated by the formula;

[0071] : water pump power and fan power, unit: watt , as a control variable, is the output of the subsequent optimization algorithm;

[0072] : battery average temperature, cooling liquid inlet temperature, environment temperature, unit: Celsius Among them, and are the real-time collection values of the corresponding position sensor; is the arithmetic mean value calculated in real time according to the readings of multiple temperature sensors in the battery pack, used to represent the overall thermal state of the battery pack;

[0073] : model coefficient, wherein is a dimensionless nonlinear power index, is the equivalent heat transfer coefficient, whose units are and to ensure dimensional consistency; The four parameters are not physically derived, but are determined through system identification experiments on the entire thermal management system; Specifically, in the calibration stage, a series of different power combination instructions are applied , while monitoring the temperature difference data when the system reaches steady state , finally based on these calibration data sets, the least squares method and other identification algorithms are used to fit;

[0074] Based on the calculated battery heat generation rate and battery heat dissipation rate, a dynamic evolution prediction model of the battery temperature state is established; this model is used to predict the temperature state of the next control period;

[0075] The evolution of the highest temperature is derived using the law of conservation of energy, and its model is:

[0076]

[0077] wherein, Cp is the equivalent heat capacity of the battery pack, in Joule per Kelvin (J / K), which is calibrated experimentally; Tc is the control period, which is a pre-set parameter for the controller; the rest of the parameters are defined or calculated in the previous steps;

[0078] Predicting the maximum temperature difference The evolution of the maximum temperature difference is predicted using a simplified linear state-space model, which is given by:

[0079]

[0080] In this equation, is the dimensionless temperature difference decay coefficient, and are the heat generation imbalance coefficient and the heat dissipation imbalance coefficient, respectively, both in Kelvin per Watt to ensure dimensional consistency; these three coefficients are calibrated and identified from computational fluid dynamics (CFD) simulations or extensive bench test data; during the calibration process, by analyzing the temperature field distribution data under different heat flow input conditions, these three empirical parameters representing the inherent thermal characteristics of the system can be fitted;

[0081] By using the above modeling method, a high-fidelity thermal characteristic prediction model is constructed; this model not only accurately predicts the future maximum temperature and maximum temperature difference caused by any power command, but also has a clear structure and efficient calculation, fully meeting the real-time optimization requirements of the vehicle-mounted controller; this precise prediction capability is the fundamental guarantee for the entire optimization method to achieve accurate and energy-saving control.

[0082] Embodiment 3:

[0083] S2 includes:

[0084] Setting the energy consumption optimization goal as minimizing the sum of the power of the cooling system water pump and fan;

[0085] Setting the initial safety constraint as the maximum temperature of the battery pack not exceeding the pre-set maximum temperature safety upper limit, and the maximum temperature difference of the battery pack not exceeding the pre-set maximum temperature difference consistency upper limit;

[0086] Constructing the Lagrange optimization function includes:

[0087] Based on the predicted maximum temperature and the maximum temperature safety upper limit, constructing a temperature constraint function;

[0088] Based on the predicted maximum temperature difference and the maximum temperature difference consistency upper limit, constructing a temperature difference constraint function;

[0089] Combining the energy consumption optimization goal, the temperature constraint function, and the temperature difference constraint function, constructing a Lagrange optimization function;

[0090] This embodiment is based on embodiment 1, and the construction process of the energy consumption and safety constraint optimization model in step S2 is described in detail. The abstract control target is converted into precise mathematical language;

[0091] To achieve the foregoing object, the energy consumption optimization target is defined in this embodiment. The target is set to minimize the sum of the power of the cooling system water pump and fan. This definition intuitively and effectively represents the main power consumption of the thermal management system. In mathematics, the objective function is defined as:

[0092]

[0093] where is the control variable vector containing the water pump power and fan power;

[0094] At the same time, the initial safety constraint is defined in this embodiment. The constraint is set to the maximum temperature of the battery pack not exceeding the preset maximum temperature safety upper limit, and the maximum temperature difference of the battery pack not exceeding the preset maximum temperature difference consistency upper limit. The two upper limit values and are the design specifications and industry safety regulations of the battery, which together constitute the core boundary of the safe operation of the battery. For example, for a ternary lithium battery, the initial maximum temperature safety upper limit may be set to , and the maximum temperature difference consistency upper limit may be set to ;

[0095] Based on the above target and constraint, the core step of this embodiment is to construct the Lagrange optimization function. This process converts an optimization problem with inequality constraints into an equivalent unconstrained problem for solving. The specific construction process is as follows:

[0096] Construct the temperature constraint function ; this function is based on the predicted maximum temperature output by the dynamic evolution prediction model established in embodiment 2 (formula ) and the maximum temperature safety upper limit , and its purpose is to convert the temperature safety requirement into a standard inequality constraint form. Its expression is:

[0097]

[0098] When the inequality holds, it means that the temperature is within the safe range;

[0099] Construct the temperature difference constraint function ; this function is based on the predicted maximum temperature difference Upper limit of maximum temperature difference consistency The purpose is to convert the battery pack internal temperature consistency requirement into a standard inequality constraint form; its expression is:

[0100]

[0101] When the inequality is established, it means that the battery pack temperature consistency meets the requirements;

[0102] Combine the energy consumption optimization goal, temperature constraint function and temperature difference constraint function to build the final Lagrange optimization function ; The function integrates all the above elements into a unified mathematical framework:

[0103]

[0104] Where, is the Lagrange multiplier vector, in order to ensure dimensional consistency, the multiplier and The unit should be power / temperature, such as W / K; They represent the marginal energy consumption cost that the system needs to pay to maintain the constraint boundary when the corresponding constraint is activated; Its components and ; These multipliers are not preset parameters, but result variables calculated together with the optimal power during the optimization solution process;

[0105] The embodiment accurately converts the engineering target energy saving and the safety boundary temperature and temperature difference limit into mathematical objective function and inequality constraint, and uses the Lagrange multiplier method to build a unified optimization function, thereby laying a solid foundation for subsequent numerical solution; This systematic mathematical modeling method ensures that the optimization problem has a solution and the solution is optimal, achieving the coordinated optimization of energy consumption and safety, two core indicators.

[0106] Example 4:

[0107] The battery aging state parameters include the battery health state and the equivalent cycle number estimated by the battery management system;

[0108] The calculation of the battery aging factor includes:

[0109] Based on the battery health state and the equivalent cycle number, the battery aging factor is calculated through a preset aging factor model;

[0110] Generating the corrected dynamic safety limit value includes:

[0111] Based on the calculated battery aging factor, the highest temperature safety upper limit preset in the initial safety constraint is corrected in a proportional adjustment manner to generate the corrected highest temperature limit value;

[0112] Based on the calculated battery aging factor, the upper limit of the maximum temperature difference consistency preset in the initial safety constraint is corrected in a proportional adjustment manner to generate a corrected maximum temperature difference limit value;

[0113] The embodiment on the basis of embodiment 1, the battery aging state monitoring and constraint dynamic correction part in step S3 is implemented, which is the key to realize the full life cycle adaptive control of the application;

[0114] In this embodiment, the battery aging state parameters are specifically the battery state of health SOH and the equivalent cycle number Nc estimated by the battery management system BMS; the battery state of health SOH( ) refers to the percentage of the current battery relative to the new battery that can store and release the amount of electricity, which is usually the estimation result of the BMS based on the online identification algorithm such as extended Kalman filter; the equivalent cycle number refers to the standard cycle number converted from different depths and different rates of charge and discharge process, which is calculated by the BMS through ampere-hour integration method combined with experience weight accumulation;

[0115] The embodiment calculates the battery aging factor through a preset aging factor model; the purpose of this step is to fuse the information of multiple aging dimensions into a single, normalized index; the expression of the model is:

[0116]

[0117] Among them, : aging factor, dimensionless, calculated by the formula;

[0118] : battery state of health and equivalent cycle number, both are real-time estimation values of the BMS;

[0119] : design reference cycle life, unit: times, is a design parameter provided by the battery manufacturer;

[0120] : weight coefficient, dimensionless, is an adjustable parameter set according to the specific battery degradation mechanism and accelerated aging test data, usually satisfies ;

[0121] The most critical step of the embodiment is to generate the revised dynamic safety limit; the embodiment adopts a proportional adjustment method to revise the preset limit in the initial safety constraint, which is clear in principle and easy to implement in engineering; it should be noted that the linear weighting model is an engineering simplification of the complex battery aging physical process; the purpose is to obtain a quantitative index that can macroscopically represent the battery aging degree at a lower calculation cost, thereby meeting the demand of real-time dynamic adjustment of the control strategy of the vehicle-mounted controller; the model considers two key aging dimensions of capacity attenuation (SOH) and cycle wear (Nc), and is suitable for degradation scenarios dominated by cycle life;

[0122] Revision of the maximum temperature safety upper limit:

[0123]

[0124] The revised maximum temperature limit is generated by subtracting a quantity proportional to the aging factor from the initial limit ; wherein, is a dimensionless revision sensitivity coefficient; the coefficient is determined based on accelerated aging test data, specifically, by conducting cycle aging tests on the battery at different constant ambient temperatures , a series of data points about temperature and battery cycle life are obtained, and the degree of influence of temperature on life is determined through regression analysis and other methods, thereby calibrating ; the reference ambient temperature , for example 25°C, is an industry standard or design specification; the linear proportional adjustment method here is an effective engineering approximation, which assumes that within a certain range, the amount of temperature safety upper limit reduction is proportional to the aging factor, and this relationship is guaranteed to be effective through the accurate calibration of the revision sensitivity coefficient ; the value of is usually between 0.1 and 0.5 to reflect the sensitivity of different types of batteries to temperature;

[0125] Revision of the maximum temperature difference consistency upper limit:

[0126]

[0127] The generation logic of the revised maximum temperature difference limit is similar, and the revision amount is directly proportional to the initial limit and the aging factor ; is the corresponding dimensionless revision sensitivity coefficient, and its calibration method is similar to , by analyzing different constant temperature differences Based on the battery pack aging experimental data, the influence of temperature difference on the inconsistency of battery pack life attenuation is quantified to determine its value; similar to the temperature limit correction, the linear model simplifies the complex relationship between temperature difference and life, and the parameter calibration driven by experimental data ensures its engineering practicability;

[0128] This embodiment successfully links the long-term aging state of the battery to the short-term optimization control target by introducing a quantifiable aging factor and establishing a proportional correction mechanism for dynamic safety limits; This feedforward dynamic correction strategy enables the thermal management system to foresee the higher safety risk of the aging battery and actively adopt a more conservative operating strategy; Compared with the traditional method of using fixed safety limits throughout the life cycle, the present invention can significantly delay the aging process of the battery, thereby maximizing the economic value and service life of the battery under the premise of ensuring safety;

[0129] To ensure the robustness of the model, boundary checks need to be performed on the calculation results in actual applications; for example, set an upper limit for the calculation result of the aging factor , such as , to prevent extreme values caused by sensor abnormalities or model extrapolation; at the same time, set reasonable lower limits for the corrected dynamic safety limits and , for example, stipulate that cannot be lower than a certain preset minimum operating temperature, such as , cannot be lower than a certain minimum allowable temperature difference, such as , to avoid physically meaningless or unmet constraint conditions and ensure that the optimization solver can run stably under all working conditions;

[0130] The model proposed in this embodiment is a simplified model constructed for real-time optimization, effectively balancing accuracy and computational efficiency; In future research, more aging characteristic parameters can be introduced to construct a nonlinear dynamic safety boundary model to further improve the adaptability and accuracy of the model under complex working conditions throughout the life cycle; It should be noted that the linear proportional adjustment model used currently is an effective engineering approximation within the main health interval of the battery, and its advantages lie in simple calculation, easy calibration and online implementation; For the strong nonlinear degradation behavior at the end of the battery life, this model may have some deviations;

[0131] In addition, the current model mainly focuses on aging caused by capacity attenuation and cycle number, and for special aging scenarios dominated by other specific factors, such as long-term low-temperature storage, high-rate lithium precipitation, etc., additional state parameters may need to be introduced to compensate for the calculation of the aging factor to further improve the generalization ability of the model.

[0132] Embodiment 5

[0133] solving includes:

[0134] In each control cycle, the Karush-Kuhn-Tucker (KKT) optimality condition corresponding to the updated Lagrangian optimization function is solved to obtain the optimal power allocation;

[0135] The optimal power allocation is output as an optimal cooling system power instruction to the water pump and fan driver of the cooling system;

[0136] This embodiment is based on Embodiment 1 and describes in detail the final solving step in step S3, explaining how to convert the constructed mathematical model into specific control instructions;

[0137] In each control cycle, the Karush-Kuhn-Tucker (KKT) optimality condition corresponding to the updated Lagrangian optimization function is solved to obtain the optimal power allocation; the KKT optimality condition is a first-order necessary condition for the optimal solution of a constrained optimization problem, and is a standard mathematical method for solving the Lagrangian function; the core logic is reflected in the following aspects:

[0138] Gradient condition : At the optimal solution, the gradient of the Lagrangian function with respect to the control variable, i.e. power, must be zero;

[0139] Complementary slackness condition : This is the key to achieving energy saving; the internal logic is that if a constraint is relaxed, i.e. , the temperature is far below the limit, then the corresponding Lagrange multiplier must be zero, at which time the system will have the sole goal of minimizing energy consumption; only when a constraint is activated, i.e. , the temperature is about to reach the limit, the corresponding multiplier is allowed to be positive, thus allowing the system to increase energy consumption to strictly meet the constraint;

[0140] Primal feasibility ( ) and dual feasibility ( ) ensure that the solution is within the safety boundary;

[0141] In a vehicle-mounted controller, this set of nonlinear equations is usually solved in real time using efficient numerical iterative algorithms such as Sequential Quadratic Programming (SQP) or Interior Point Method (IPM);

[0142] After solving the KKT condition, the result is the optimal power allocation The optimal power distribution is taken as an optimal cooling system power instruction and output to the water pump and fan driver of the cooling system. After receiving the specific power value, the driver accurately controls the motor operation through pulse width modulation (PWM) and other methods to realize physical execution of the instruction.

[0143] The state perception-model prediction-constraint correction-optimization solution-control execution process is continuously cycled in each control period, for example, 1 second, to form a complete closed-loop control system.

[0144] The embodiment successfully converts the complex optimization theory into an engineering-achievable closed-loop control logic. By efficiently solving the KKT condition in each control period, the method can ensure that the power output of the cooling system is always the optimal solution under the current working condition and the battery health state, that is, the energy consumption is the lowest on the premise of strictly complying with the dynamic safety boundary. The real-time and continuous optimization capability enables the system to respond to any changes in the external environment and internal state in real time, thereby practically converting the theoretical energy-saving potential and life-prolonging effect into stable benefits in actual operation.

[0145] The above is only a preferred embodiment of the present application, and is not intended to limit the protection scope of the present application; any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0146] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A battery thermal management system optimization method, characterized by, The specific steps include: S1, acquiring the operating state parameters and the battery aging state parameters of the battery pack; based on the operating state parameters, constructing a thermal characteristic prediction model to generate a predicted maximum temperature and a predicted maximum temperature difference; S2, setting an energy consumption optimization target and initial safety constraints; combining the predicted maximum temperature and the predicted maximum temperature difference, constructing a Lagrange optimization function considering the initial safety constraints; S3, based on the battery aging state parameters, calculating a battery aging factor; according to the battery aging factor, dynamically modifying the initial safety constraints to generate modified dynamic safety limits; using the modified dynamic safety limits to update the Lagrange optimization function, and solving the updated Lagrange optimization function to output an optimal cooling system power instruction; Generating the modified dynamic safety limits includes: Based on the calculated battery aging factor, the maximum temperature safety upper limit preset in the initial safety constraints is modified in a proportional adjustment manner to generate a modified maximum temperature limit; the modification formula of the maximum temperature safety upper limit is: wherein, is the modified maximum temperature limit; is the initial maximum temperature limit; is a dimensionless correction sensitivity factor; is the ambient temperature; is the aging factor; Based on the calculated battery aging factor, the maximum temperature difference consistency upper limit preset in the initial safety constraints is modified in a proportional adjustment manner to generate a modified maximum temperature difference limit; the modification formula of the maximum temperature difference consistency upper limit is: wherein, is the modified maximum temperature difference limit; is the initial maximum temperature difference limit; is the corresponding dimensionless modified sensitivity coefficient.

2. The method of claim 1, wherein, S1 includes: Real-time acquisition of the charging and discharging current and the cell temperature of the battery pack; and real-time acquisition of the external environment temperature and the cooling liquid inlet temperature; Based on the acquired charging and discharging current and the cell temperature, the battery heat generation rate is calculated by using the modified Bernard heat model; Based on the acquired cell temperature, cooling liquid inlet temperature and external environment temperature, the battery heat dissipation rate is calculated by using the heat dissipation model determined through system identification experiment.

3. The method of claim 2, wherein, S1 further includes: Based on the calculated battery heat generation rate and battery heat dissipation rate, a dynamic evolution prediction model of the battery temperature state is established; The dynamic evolution prediction model generates a predicted maximum temperature by using energy conservation deduction, and generates a predicted maximum temperature difference by using a simplified linear state space model.

4. The method of claim 1, wherein S2 It includes: The energy consumption optimization target is set as the sum of the power of the cooling system water pump and fan being minimized; The initial safety constraints are set as the maximum temperature of the battery pack not exceeding the preset maximum temperature safety upper limit, and the maximum temperature difference of the battery pack not exceeding the preset maximum temperature difference consistency upper limit.

5. The method of claim 4, wherein, The construction of the Lagrange optimization function includes: Based on the predicted maximum temperature and the maximum temperature safety upper limit, a temperature constraint function is constructed; Based on the predicted maximum temperature difference and the maximum temperature difference consistency upper limit, a temperature difference constraint function is constructed; Combining the energy consumption optimization target, the temperature constraint function and the temperature difference constraint function, the Lagrange optimization function is constructed.

6. The method of claim 1, wherein, The battery aging state parameters include the battery health state and the equivalent cycle number estimated by the battery management system; The calculation of the battery aging factor includes: Based on the battery health state and the equivalent cycle number, the battery aging factor is calculated through a preset aging factor model.

7. The method of claim 1, wherein, The solution includes: In each control period, the Karush-Kuhn-Tucker optimality condition corresponding to the updated Lagrange optimization function is solved to obtain the optimal power distribution; The optimal power distribution is taken as the optimal cooling system power instruction and output to the water pump and fan driver of the cooling system.

Images