A power battery cooling control method, storage medium, electronic device and vehicle

Abstract

This application provides a power battery cooling control method, storage medium, electronic device, and vehicle. The method includes: acquiring measurement signals; obtaining the instantaneous heat generation rate of the battery based on battery temperature, battery current, and a preset battery heat load estimation model; obtaining the average dryness of the cold plate and the boiling front position based on the coolant pump speed, cold plate outlet dryness, and a preset cold plate two-phase flow dynamic model; calculating the critical heat flux density margin of the cold plate based on the instantaneous heat generation rate of the battery; calculating the boiling probability using a pre-trained boiling prediction model based on the cold plate outlet dryness, average cold plate dryness, boiling front position, and critical heat flux density margin; and calculating a pump speed control command using a multi-objective model predictive controller based on the average cold plate dryness, boiling front position, and boiling probability. The pump speed control command is used to change the coolant flow rate to adjust the cooling speed of the power battery. This solution can perceive the distribution of the boiling region inside the cold plate in real time and predict the risk of burning out.

Description

Technical Field

[0001] This application relates to the field of thermal management technology during electric vehicle battery charging, specifically to a power battery cooling control method, storage medium, electronic equipment, and vehicle. Background Technology

[0002] During fast charging of electric vehicles, two-phase flow boiling cooling is a key heat dissipation method. However, once the heat flux exceeds a critical value, nucleation boiling can abruptly transform into film boiling (burning out), triggering battery thermal runaway. Existing control strategies rely solely on outlet signal feedback to control the cooling process, failing to detect the distribution of boiling regions inside the cold plate in real time or predict the risk of burning out, thus still posing safety hazards. Summary of the Invention

[0003] The technical problem this application aims to solve is that when existing vehicle battery thermal management systems use two-phase flow boiling cooling technology to cool the battery, they lack the ability to perceive and predict the key internal states of the boiling cooling process in real time. As a result, they cannot dynamically optimize heat dissipation efficiency and system energy consumption while ensuring safety. Therefore, this application provides a power battery cooling control method, storage medium, electronic equipment, and vehicle.

[0004] In a first aspect, the technical solution of this application provides a power battery cooling control method, including: Acquire measurement signals, including battery temperature, battery current, coolant pump speed, and cold plate outlet dryness; The instantaneous heat generation rate of the battery is obtained based on the battery temperature, battery current, and preset battery heat load estimation model. Based on the coolant pump speed, cold plate outlet dryness, and the two-phase flow dynamic model of the pre-installed cold plate, the average dryness of the cold plate and the boiling front position are obtained. Calculate the critical heat flux density margin of the cold plate based on the instantaneous heat generation rate of the battery; The boiling probability is calculated using a pre-trained boiling prediction model based on the cold plate outlet dryness, cold plate average dryness, boiling front position, and critical heat flux margin. Based on the average dryness of the cold plate, the position of the boiling front, and the boiling probability, a multi-objective model predicts the pump speed control command to calculate the pump speed control command. The pump speed control command is used to change the coolant flow rate to adjust the cooling rate of the power battery.

[0005] In some solutions for power battery cooling control methods, the step of obtaining the instantaneous heat generation rate of the battery based on battery temperature, battery current, and a preset battery heat load estimation model is expressed as follows: ; in, Indicates the instantaneous heat generation rate of the battery. This indicates the battery current. This indicates the ohmic internal resistance of the battery. This indicates the battery temperature. It is the entropy heat coefficient.

[0006] In some solutions for power battery cooling control methods, the cold plate is divided into two regions along the refrigerant flow direction: a single-phase liquid region from the inlet to the boiling front, and a two-phase boiling region from the boiling front to the outlet; in the step of obtaining the average dryness of the cold plate and the boiling front position based on the coolant pump speed, the dryness of the cold plate outlet, and a pre-set two-phase flow dynamic model of the cold plate, the two-phase flow dynamic model of the cold plate is expressed as: ; ; in, This indicates the average dryness of the cold-rolled steel plate. The boiling front position; and These represent the densities of the gas and liquid phases, respectively. Indicates the latent heat of vaporization; Indicates the cross-sectional area of ​​the cold plate flow channel; L represents the length of the two-phase region, and L represents the total length of the cold plate. This represents the mass flow rate of the refrigerant, which is calculated from the pump characteristic curve based on the coolant pump speed. Indicates the dryness of the cold plate inlet; This represents the amount of heat absorbed by the battery in the two-phase region. Assuming the heat flux density is uniform, then... ; This represents the heat absorbed by the single-phase region. ; This represents the change in specific enthalpy of a single-phase liquid.

[0007] In some solutions for power battery cooling control methods, the critical heat flux density margin is calculated in the step of calculating the critical heat flux density margin of the cold plate based on the instantaneous heat generation rate of the battery, using the following method: ; ; ; in, This indicates the critical heat flux density margin. Indicates surface tension. Indicates the dryness of the cold plate outlet. and Represents an empirical constant. This represents the current maximum heat flux density. A critical heat flux density margin greater than 0 indicates safety, while a margin less than 0 indicates that the material has burned out.

[0008] In some solutions, the power battery cooling control method includes the following step: calculating the boiling probability using a pre-trained boiling prediction model based on the cold plate outlet dryness, average cold plate dryness, boiling front position, and critical heat flux margin. The dryness of the cold plate outlet is sampled using a sliding window, and the time-domain statistical features and frequency-domain features of the signal within the window are extracted. The time-domain statistical features and the frequency-domain features are combined with the average dryness of the cold plate, the boiling front position, and the critical heat flux margin to form a multi-dimensional feature vector; The multidimensional feature vector is input into the pre-trained boiling prediction model, and the boiling probability is output.

[0009] In some solutions, the power battery cooling control method uses boiling probability as the probability of film boiling occurring within a set time period; the set time is less than 10 seconds.

[0010] In some solutions, the power battery cooling control method uses a support vector machine classifier as the boiling prediction model, which is expressed as follows: ; in, The boiling probability is represented by H and B, which are pre-fitted constants. This represents the classification score of the support vector machine classifier.

[0011] In some solutions for power battery cooling control methods, the step of calculating pump speed control commands based on the average dryness of the cold plate, the position of the boiling front, and the boiling probability using a multi-objective model predictive controller includes: Control Model: ; ; in, This indicates the average dryness of the cold-rolled steel plate. This indicates the average wall temperature of the cold-rolled steel plate. Indicates the coolant pump speed. Indicates the instantaneous heat generation rate of the battery; a ij , b i , d i These represent the model coefficients, where i and j are both 1 or 2. and Indicates the period number; Objective function: ; in, Indicates the predicted future number The highest battery temperature of the step, Indicates the target battery temperature; α and β This represents the weighting coefficient.

[0012] In some of the power battery cooling control methods described, the multi-objective model predictive controller further includes: Constraints: ,in ε This is a safety threshold; ; in, Indicates the predicted future number The critical heat flux margin of the step, and This indicates the lower and upper threshold values ​​for the coolant pump speed.

[0013] The power battery cooling control method described in some solutions further includes: The pump speed control command is represented as a control sequence. , with the first value As the optimal control value for the current cycle; When the boiling probability is less than the set threshold, the system enters the performance optimization mode. When the boiling probability is greater than or equal to the set threshold, the system enters a safety-first mode, and a penalty term is added to the objective function. , γ The penalty coefficient is increased; the safety threshold is raised, and the weighting is adjusted. α reduce.

[0014] Secondly, the present application provides a power battery cooling control device, comprising: The signal acquisition module is used to acquire measurement signals, including battery temperature, battery current, coolant pump speed, and cold plate outlet dryness. The first calculation module is used to obtain the instantaneous heat generation rate of the battery based on the battery temperature, battery current and the preset battery heat load estimation model. The second calculation module is used to obtain the average dryness of the cold plate and the boiling front position based on the coolant pump speed, the dryness of the cold plate outlet and the pre-set two-phase flow dynamic model of the cold plate. The first calculation module is used to calculate the critical heat flux density margin of the cold plate based on the instantaneous heat generation rate of the battery. The fourth calculation module is used to calculate the boiling probability based on the cold plate outlet dryness, the average dryness of the cold plate, the boiling front position, and the critical heat flux margin, using a pre-trained boiling prediction model. The control output module is used to predict and calculate the pump speed control command based on the average dryness of the cold plate, the position of the boiling front, and the boiling probability using a multi-objective model. The pump speed control command is used to change the coolant flow rate to adjust the cooling speed of the power battery.

[0015] Thirdly, the present application provides a computer-readable storage medium storing program information, wherein a computer reads the program information and executes the steps of the power battery cooling control method described in any one of the first aspects.

[0016] Fourthly, the present application provides a computer program product, including a computer program / instructions, which, when executed by a processor, implements the steps of the power battery cooling control method described in any of the first aspects.

[0017] Fifthly, the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the power battery cooling control method according to any one of the first aspects.

[0018] Sixthly, this application provides a vehicle including the power battery cooling control device described in the second aspect, or the computer-readable storage medium described in the third aspect, or the computer program product described in the fourth aspect, or the electronic device described in the fifth aspect.

[0019] The technical solution provided in this application has the following technical effects compared with the prior art: This application's solution, through a battery thermal load estimation model and a cold plate two-phase flow dynamic model, utilizes real-time measurement signals of battery temperature, battery current, coolant pump speed, and cold plate outlet dryness to calculate unmeasurable states such as the boiling front position and average dryness inside the cold plate, solving the problem of existing technologies being unable to perceive the distribution of boiling regions inside the cold plate. Based on this, by calculating the critical heat flux margin of the cold plate and combining it with outlet dryness characteristics, a pre-trained boiling prediction model outputs the boiling probability, achieving early warning of burn-out risk and overcoming the lag defect of traditional outlet signal feedback control. Based on the estimated state and predicted risk probability, a multi-objective model predictive controller dynamically optimizes the pump speed control command to achieve coolant speed control, meeting the cooling requirements of the power battery. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of a two-phase flow cooling system for a power battery. Figure 2 This is a flowchart of a power battery cooling control method according to one embodiment of this application; Figure 3This is a flowchart of a power battery cooling control method according to another embodiment of this application; Figure 4 This is a structural block diagram of a power battery cooling control device according to one embodiment of this application; Figure 5 This is a schematic diagram of the structure of an electronic device that performs the power battery cooling control method according to an embodiment of this application. Detailed Implementation

[0021] The specific embodiments of this application will be further described below with reference to the accompanying drawings.

[0022] It is readily understood that, based on the technical solution of this application, various structural and implementation methods can be interchanged by those skilled in the art without altering the essential spirit of this application. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this application and should not be considered as the entirety of this application or as limitations or restrictions on the technical solution of the application.

[0023] Reference Figure 1 The system architecture of this embodiment includes a power battery 100, which integrates a microchannel cold plate 101. Driven by an electronic pump 102, a low-temperature two-phase refrigerant flows through the cold plate 101, absorbs heat from the battery, and partially vaporizes, forming a gas-liquid two-phase mixture. This mixture enters the condenser 103, where it is completely condensed into a subcooled liquid with the assistance of a fan 104, completing the cycle. To achieve intelligent control, the system is equipped with the following key sensors: a temperature and pressure sensor 105 is installed at the inlet of the cold plate 101; a first dryness sensor 106 is installed at the inlet of the cold plate 101, and a second dryness sensor 107 is installed at the outlet; multiple temperature sensors 108 are integrated inside the power battery 100; and the battery management system provides a total current I... batt and total voltage V batt Signals. All sensor data and signals are transmitted in real time to the thermal management controller 200 via the CAN bus at a fixed sampling period (e.g., 100ms). The solution proposed in this application is described in conjunction with this system architecture.

[0024] This application provides a power battery cooling control method, applied in the controller of a thermal management system, such as... Figure 2 As shown, it includes: S100: Acquire measurement signals, including battery temperature, battery current, coolant pump speed, and cold plate outlet dryness.

[0025] Specifically, battery temperature is collected by multiple temperature sensors integrated inside the power battery, and the average value of the sensor measurements (i.e., average temperature) is taken as the current battery temperature. Battery current is provided by the battery management system (BMS) and expressed as total battery current. Coolant pump speed is obtained through a pump speed sensor or by reading the feedback signal from the electronic pump via the CAN bus. Cold plate outlet dryness is measured in real time by a dryness sensor installed on the cold plate outlet pipeline. All measurement signals are transmitted to the thermal management controller via the CAN bus at a fixed sampling period (e.g., 100ms).

[0026] S200: The instantaneous heat generation rate of the battery is obtained based on the battery temperature, battery current and the preset battery heat load estimation model.

[0027] Specifically, the battery temperature and battery current obtained in step S100 are input into a pre-established battery heat load estimation model. This model describes the quantitative relationship between the battery's instantaneous heat generation rate and its temperature and current. The model is pre-established through offline experiments, and its parameters are stored in the controller in the form of a MAP table. For example, a MAP table of the battery's ohmic internal resistance and entropy thermal coefficient is obtained through a hybrid pulse power characteristic experiment. The experiment is conducted at different SOCs (0-100%, step size 10%) and different temperatures (-20℃ to 50℃, step size 10℃), recording pulse response data, constructing a three-dimensional MAP table, and storing it in the controller. During this step, the controller calculates the current instantaneous battery heat generation rate based on the current battery temperature and current using the battery heat load estimation model. The instantaneous battery heat generation rate reflects the heating intensity of the power battery under the current operating conditions and serves as the basis for subsequent cold plate heat load calculations.

[0028] S300: Based on the coolant pump speed, cold plate outlet dryness, and the dynamic model of the two-phase flow of the pre-installed cold plate, the average dryness of the cold plate and the boiling front position are obtained.

[0029] The two-phase flow dynamic model for the cold plate is used to describe the dynamic characteristics of the refrigerant flow and boiling process within the cold plate. In some embodiments, this model is pre-established through offline experiments and characterized in the form of differential equations or state-space equations. Its parameters are pre-calibrated based on the geometry of the cold plate and the physical properties of the refrigerant. For example, a two-phase flow cold plate experimental rig is built, and the system is run under different heat loads (simulating battery heat generation) and flow rates, recording response data such as inlet temperature, pressure, outlet dryness fraction, and wall temperature. The unknown parameters in the two-phase flow dynamic model of the cold plate are fitted using the nonlinear least squares method to minimize the error between the model output and the experimental data.

[0030] During this step, the controller estimates two key states inside the cold plate that cannot be directly measured in real time, based on the current coolant pump speed (converted to mass flow rate) and the dryness of the cold plate outlet, using the aforementioned two-phase flow dynamic model: the average dryness of the cold plate and the boiling front position. The average dryness of the cold plate reflects the average proportion of refrigerant vapor in the two-phase region; the boiling front position reflects the position where the refrigerant begins to boil from a single-phase liquid, and is an important physical indicator characterizing the risk of burn-out.

[0031] S400: Calculate the critical heat flux density margin of the cold plate based on the instantaneous heat generation rate of the battery.

[0032] Specifically, firstly, the maximum heat flux density that the cold plate currently withstands is calculated based on the battery's instantaneous heat generation rate and the cold plate's heat exchange area. Simultaneously, the critical heat flux density (i.e., the critical value at which film boiling occurs) of the cold plate under this operating condition is calculated based on the cold plate outlet dryness. Then, the critical heat flux density margin is calculated, which is the relative difference between the critical heat flux density and the current maximum heat flux density. A margin value greater than 0 indicates that the current state is safe; a smaller margin indicates that the cold plate is closer to the risk of burning out. The critical heat flux density margin quantifies how much safety margin the cold plate has left before film boiling (burning out).

[0033] S500: The boiling probability is calculated using a pre-trained boiling prediction model based on the cold plate outlet dryness, average cold plate dryness, boiling front position, and critical heat flux margin.

[0034] In this step, the cold plate outlet dryness obtained in step S100, the average cold plate dryness and boiling front position obtained in step S300, and the critical heat flux margin obtained in step S400 are jointly input into the pre-trained boiling prediction model. The boiling prediction model is pre-trained using offline experimental data. Its input is the aforementioned multidimensional state information, and its output is the probability value (i.e., boiling probability) of film boiling occurring within a preset future timeframe. This model can capture the precursor features of the transition from normal nucleus boiling to film boiling, achieving an advanced assessment of the risk of burn-out. Specifically, in each control cycle, the multidimensional state information of the current moment is input, and the model outputs a continuous probability value between 0 and 1. This probability value directly characterizes the risk level of the system: a lower probability value indicates greater safety, while a higher probability value indicates a greater risk of burn-out.

[0035] S600: Based on the average dryness of the cold plate, the position of the boiling front, and the boiling probability, the pump speed control command is calculated by a multi-objective model predictive controller; the pump speed control command is used to change the coolant flow rate to adjust the cooling speed of the power battery.

[0036] Based on the average dryness of the cold plate and the boiling front position obtained in step S300, and the boiling probability obtained in step S500, the pump speed control command is calculated by a multi-objective model predictive controller. Specifically, the multi-objective model predictive controller uses the average dryness of the cold plate and the cold plate wall temperature as state variables, the pump speed control command as the control variable, and the instantaneous heat generation rate of the battery as a disturbance input to establish a predictive model. In each control cycle, the controller predicts the system state within a finite time domain and solves a constrained multi-objective optimization problem. Optimization objectives may include battery temperature tracking accuracy and pump energy consumption, while constraints include the critical heat flux density margin safety lower limit and actuator physical limitations. The pump speed control command is sent to the electronic coolant pump to change the coolant flow rate to adjust the cooling rate of the power battery. After the control command is executed, the system state changes, and the new measurement data is returned to step S100, forming a closed-loop feedback control.

[0037] The above-described embodiment, through a battery thermal load estimation model and a cold plate two-phase flow dynamic model, utilizes real-time measurement signals of battery temperature, battery current, coolant pump speed, and cold plate outlet dryness to calculate unmeasurable states such as the boiling front position and average dryness inside the cold plate, solving the problem that existing technologies cannot perceive the distribution of boiling regions inside the cold plate. Based on this, by calculating the critical heat flux margin of the cold plate and combining it with outlet dryness characteristics, a pre-trained boiling prediction model outputs the boiling probability, achieving early warning of burn-out risk and overcoming the lag defect of traditional outlet signal feedback control. Based on the estimated state and predicted risk probability, a multi-objective model predictive controller dynamically optimizes the pump speed control command to achieve coolant speed control and meet the cooling requirements of the power battery.

[0038] Preferably, in step S200, the step of obtaining the instantaneous heat generation rate of the battery based on the battery temperature, battery current, and a preset battery heat load estimation model is expressed as follows: ; in, This is expressed as the instantaneous heat generation rate of the battery. This is expressed as the battery current. This is expressed as the battery's internal resistance in ohms. This is expressed as the battery temperature. The entropy coefficient is the partial derivative of the battery's open-circuit voltage with respect to temperature. Both can be obtained from the offline calibrated "SOC-Temperature-R0" MAP table, and are determined in real-time by the controller based on the current SOC and temperature. Table lookup update. Battery instantaneous heat generation rate in this application. The model consists of two parts: Joule heat and entropy heat. It employs a first-order RC equivalent circuit model combined with the entropy heat coefficient for calculation. Joule heat reflects the irreversible heat generated by the battery's internal resistance, while entropy heat reflects the reversible heat generated by entropy change during battery charging and discharging. Especially under fast charging conditions, the contribution of entropy heat to the total battery heat generation rate is not negligible. Compared to traditional methods that only consider Joule heat, this model significantly improves the accuracy of heat generation rate calculation.

[0039] Furthermore, the cold plate in the S300 is divided into two regions along the refrigerant flow direction: a single-phase liquid region from the inlet to the boiling front, and a two-phase boiling region from the boiling front to the outlet. To balance real-time performance and accuracy, this scheme uses a moving boundary lumped parameter model to describe the boiling process within the cold plate. This model divides the cold plate into two regions along the refrigerant flow direction. Single-phase liquid region: from the inlet to the boiling initiation point (i.e., the boiling front L). b The refrigerant is a subcooled liquid, and its temperature gradually rises to the saturation temperature T. sat Two-phase boiling zone: from L b At the outlet, the refrigerant is in a two-phase state, and the dryness fraction increases linearly from 0 to x. out To describe the boiling state of the two-phase flow within the cold plate, two state variables are introduced.

[0040] Average dryness fraction x avg The average dryness fraction within the two-phase region is defined as follows: ; boiling front position L b The length of the single-phase zone reflects the extent to which the boiling zone moves towards the outlet. A smaller value means that the single-phase zone is shortened and the two-phase zone is extended towards the inlet. It is an important physical indicator of the risk of burn-out.

[0041] In the step of obtaining the average dryness of the cold plate and the boiling front position based on the coolant pump speed, the dryness of the cold plate outlet, and the pre-set two-phase flow dynamic model of the cold plate, the two-phase flow dynamic model of the cold plate can be derived according to the laws of mass conservation and energy conservation. The two-phase flow dynamic model of the cold plate is expressed as follows: ; ; in, The average dryness of the cold-rolled plate. The boiling front position; and These are the densities of the gas and liquid phases, respectively. Latent heat of vaporization; This refers to the cross-sectional area of ​​the cold plate flow channel; L is the length of the two-phase region, and L is the total length of the cold plate. The mass flow rate of the refrigerant is calculated from the pump characteristic curve based on the coolant pump speed; x in Dryness of the cold plate inlet; Let the heat absorbed by the battery in the two-phase region be assuming a uniform heat flux density. ; The heat absorbed by the single-phase region ; This represents the specific enthalpy change of a single-phase liquid.

[0042] In the above model, average dryness With boiling front position Mutual coupling: The length of the two-phase region is determined, which in turn affects the rate of change of the average dryness; and The changes in these two state variables reflect the energy balance within the two-phase region. Simultaneously estimating both state variables maintains the physical consistency of the model, avoiding non-physical estimation results (such as a mismatch between the length of the two-phase region and the average dryness), thereby significantly improving the accuracy and robustness of the state estimation and providing a reliable basis for subsequent early warning and control.

[0043] Furthermore, in step S400, the critical heat flux density margin of the cold plate is calculated based on the instantaneous heat generation rate of the battery, and the critical heat flux density margin is obtained in the following manner: ; ; ; in, To describe the critical heat flux density margin, For surface tension, For the dryness of the cold plate outlet, and It is an empirical constant. The current maximum heat flux density (CHF) is used. A critical heat flux density margin greater than 0 indicates safety, while a margin less than 0 indicates that burn-out has occurred. The critical heat flux density (CHF) is closely related to the outlet dryness and mass flow rate. This embodiment uses the SG correlation to represent it. This scheme calculates the critical heat flux density of the cold plate using the SG correlation and combines it with the real-time measured outlet dryness and the real-time calculated battery heat generation rate to dynamically obtain the current safety margin. This margin value quantifies the risk of burn-out, providing a clear safety boundary for the control system. Its dynamic characteristics enable the system to track the impact of changes in operating conditions on the safety status in real time.

[0044] Further preferably, the step in S500 of calculating the boiling probability based on the cold plate outlet dryness, the average dryness of the cold plate, the boiling front position, and the critical heat flux margin using a pre-trained boiling prediction model includes: performing sliding window sampling on the cold plate outlet dryness, extracting the time-domain statistical features and frequency-domain features of the signal within the window; combining the time-domain statistical features and the frequency-domain features with the average dryness of the cold plate, the boiling front position, and the critical heat flux margin to form a multi-dimensional feature vector; and inputting the multi-dimensional feature vector into the pre-trained boiling prediction model to output the boiling probability.

[0045] Specifically, the dryness signal x at the cold plate outlet out Perform sliding window sampling, with a window length of 10 seconds. Calculate the following characteristics of the signal within the window: variance: This reflects the fluctuation range of dryness. kurtosis: This reflects the sharpness of the signal distribution; Energy in a specific frequency band: Perform a Fast Fourier Transform on the window signal to calculate the energy proportion within the 0.5-2Hz frequency band, expressed as E. 0.5-2Hz express.

[0046] Introducing the critical heat flux margin (CHF) of the S200 output margin Average dryness fraction x avg and boiling front position L b Together, they form a state feature containing a 7-dimensional feature vector: ; Among them, the boiling front position L b It directly reflects the degree to which the boiling zone moves towards the outlet. A decrease in its value means that the single-phase zone is shortened and the two-phase zone is expanded towards the inlet. It is an important physical indicator of the risk of burn-out. The rate of change of CHF margin reflects the trend of risk changes.

[0047] In this scheme, before film boiling (burn-out) occurs, the outlet dryness signal exhibits specific dynamic change patterns, such as increased fluctuations and an increase in high-frequency components. This scheme effectively captures these precursory features of burn-out by extracting the time-domain statistical characteristics (such as variance and kurtosis) and frequency-domain characteristics (such as the energy proportion of specific frequency bands) of the outlet dryness signal, enabling the boiling prediction model to identify risk signals in advance before burn-out occurs. In addition, this scheme integrates physical state parameters such as the average dryness of the cold plate, the position of the boiling front, and the critical heat flux margin to form a multi-dimensional feature vector. By employing multi-information fusion, the prediction model can simultaneously utilize the dynamic characteristics of the signal and the physical state of the system, improving the accuracy of burn-out risk prediction.

[0048] Furthermore, the boiling probability is the probability of film boiling occurring within a set time period in the future; the set time is less than 10 seconds. This solution can output the probability value of film boiling occurring within a future preset time window (e.g., 5 seconds in the future). This solution can achieve early warning, enabling the control system to obtain risk information before burn-out occurs, providing a time window for proactive intervention, and overcoming the fundamental defect of traditional feedback control's "post-event response".

[0049] More preferably, the boiling prediction model is a support vector machine classifier, which is represented as: ; in, Let H and B be the boiling probability, and H and B be pre-fitted constants. This represents the classification score of the Support Vector Machine (SVM) classifier. This scheme uses an SVM classifier to output the result indicating whether film boiling will occur within the next 5 seconds. The training data comes from critical heat flux experiments, collecting dryness signals and corresponding CHF values ​​throughout the entire process from normal boiling to burn-off. margin x avg and L b Data is used, with samples taken 5 seconds before a crisis occurs designated as positive samples. The SVM classifier employs a radial basis function kernel function, and two key hyperparameters—the penalty factor C and the kernel parameter γ—are optimized through cross-validation to improve the classifier's predictive performance. The trained model is deployed on the controller. After calculating the feature vector in each control cycle, the SVM decision function is used to obtain a classification score, which is then mapped using a sigmoid function to obtain the boiling probability P of a crisis occurring within the next 5 seconds. crisis .

[0050] More preferably, in step S600, the step of calculating the pump speed control command based on the average dryness of the cold plate, the boiling front position, and the boiling probability using a multi-objective model predictive controller, wherein the multi-objective model predictive controller includes: Control Model: ; ; in, The average dryness of the cold-rolled plate. This indicates the average wall temperature of the cold-rolled steel plate. For coolant pump speed, For the instantaneous heat generation rate of the battery; a ij , b i , d i These are the model coefficients, where i and j are both 1 or 2. and Indicates the cycle number; in each control cycle k, the model predictive controller moves at x... avg (k) and T wall (k) represents the initial condition, and the simplified model is used to predict the future N. p The system status of the step.

[0051] Objective function: ; in, For the predicted future The highest battery temperature of the step, The target temperature for the battery; α and β These are the weighting coefficients.

[0052] Furthermore, the multi-objective model prediction controller also includes: Constraints: ,in ε This is a safety threshold; ; in, For the predicted future The critical heat flux margin of the step, and These are the lower and upper threshold values ​​for the coolant pump speed.

[0053] The above scheme uses the average dryness and average wall temperature of the cold plate as state variables, the pump speed control command as the control variable, and the instantaneous heat generation rate of the battery as the disturbance input to construct a discrete state-space prediction model. In each control cycle, the system state within a finite time domain is predicted using the current state as the initial condition, and a multi-objective optimization problem involving battery temperature tracking accuracy and pump energy consumption is solved. Constraints include a safety lower limit for the critical heat flux density margin and actuator physical limitations. Based on the comparison between the boiling probability and a preset threshold, the weight coefficients of the optimization objectives are dynamically adjusted: when the boiling probability is below the threshold, the system operates in efficiency optimization mode; when the boiling probability reaches or exceeds the threshold, it switches to safety priority mode, adding a constraint penalty term for the critical heat flux density margin. The first value of the optimal control sequence obtained is used as the pump speed control command output for the current cycle. This scheme achieves a quantitative balance between safety and efficiency through multi-objective optimization, intelligent adjustment of the control strategy through adaptive mode switching based on risk probability, and proactive prevention of burn-out risks through the forward-looking nature of model predictive control, significantly improving the safety, heat dissipation efficiency, and system energy efficiency of the fast charging process.

[0054] Furthermore, the method further includes: representing the pump speed control command as a control sequence. , with the first value This serves as the optimal control value for the current cycle; when the boiling probability is less than a set threshold, the system enters performance optimization mode; the set threshold can be selected as 0.25. Specifically, in performance optimization mode, the weighting coefficient... α =1 and β =0.1, safety threshold ε =0.05. When the boiling probability is greater than or equal to the set threshold, the system enters a safety-first mode, and a penalty term is added to the objective function. , γ The penalty coefficient is increased; the safety threshold is raised, and the weighting is adjusted. α Reduce. Specifically, γ =100; ε =0.05, α =0.5. In this scheme, the control commands are optimized and updated in real time through rolling time domain. The balance between safety and efficiency is quantified by using specific weight coefficients and safety thresholds. Through the penalty term under the safety priority, the system prioritizes the safety margin when the risk increases. The adaptive weight adjustment realizes the decision to make fine temperature control in low risk and proactive safety protection in high risk.

[0055] like Figures 1 to 3 As shown, this application is applied to a two-phase flow cooling circuit that includes a variable speed pump, a battery cold plate, a condenser, and connecting pipes. The controller executes the following three-layer algorithm flow in each operation cycle: The first layer involves observing the boiling state of the two-phase flow: the state observer receives real-time measurement signals from the system, including those calculated from the coolant pump speed. Inlet dryness of cold plate x in and export dryness x out Average battery temperature T cell Battery current I batt With voltage V batt and cold plate wall temperature T wall The instantaneous heat generation rate of the battery is calculated based on a battery heat load estimation model. Based on a two-phase flow dynamic model describing the flow boiling process within the cold plate, key internal states that cannot be directly measured are estimated in real time, including the average dryness fraction within the cold plate and the location of the boiling front. The critical heat flux margin is then calculated based on the estimation results. CHF margin This layer outputs a precise numerical estimate of the system's internal state at the current moment, providing a basis for subsequent early warning and control.

[0056] The second layer is a film boiling crisis early warning system: this intelligent early warning system receives the average dryness output from the first layer. x avgBoiling Front Position L b and critical heat flux margin CHF margin At the same time, the dryness of the cold plate outlet x out Real-time dynamic feature extraction is performed on the sensor signal, including variance, kurtosis, and specific frequency band energy. The extracted features are combined with the state parameters output from the first layer to form a 7-dimensional feature vector f. warn Input a pre-trained support vector machine classifier and output the probability of membrane boiling occurring within a preset time window. P crisis This layer outputs continuous probability values ​​that characterize the system's risk level, which are used to guide the dynamic adjustment of the third-layer control strategy.

[0057] The third layer, multi-objective model intelligent control: the model control receives the average dryness estimated by the first layer. x avg Boiling Front Position L b Crisis probability of the second layer output P crisis Simultaneously receive the cold plate wall temperature T wall As state feedback, the controller quantitatively predicts the system state over a finite time domain and solves a constrained multi-objective optimization problem. Optimization objectives include battery temperature tracking accuracy and pump energy consumption, while constraints include a safety lower bound for the critical heat flux density margin and actuator physical limitations. The controller then calculates the crisis probability based on the second-layer output. P crisis Based on the comparison results with the threshold (0.25), the weight coefficients of the optimization objective are dynamically adjusted (e.g., α , β and safety threshold ε To achieve an efficiency optimization mode ( P crisis <0.25) and security priority mode ( P crisis Adaptive switching with a value ≥0.25) outputs the optimal pump speed control command. u pump .

[0058] The logical relationship between the above three layers of algorithms is as follows: The first layer provides a numerical prediction of the current boiling state of the power battery cooling system. x avg , L b , CHF margin The second layer provides a probability assessment of the risk of future film boiling. P crisisThe third layer integrates both approaches, calculating the optimal control command through quantitative prediction and optimization decision-making. u pump The probability output of the second layer is used to dynamically adjust the control mode and weight configuration of the third-layer optimization problem, so as to achieve an intelligent balance between safety and efficiency. After the control command of the third layer is executed, the system state changes and the new measurement data enters the first layer again, forming a closed-loop feedback.

[0059] To more clearly illustrate the effectiveness of this application, the following specific scenarios will be used as examples: Initial state: Battery is discharging at 1C, temperature is stable, P crisis =0.1, MPC is operating in normal mode, and the pump speed is stable.

[0060] Fast charging initiated: 350kW fast charging begins, I batt Sudden rise, The number of x predicted by the first-layer algorithm increases rapidly. avg Rise, L b Decrease (boiling front moves towards inlet), CHF margin It has begun to descend.

[0061] Crisis warning: The second-layer algorithm detected x out Signal fluctuations intensify, variance Var and E 0.5-2Hz Energy increases, while CHF margin Continuously decreasing, L b Rapidly decrease, P crisis It rose from 0.1 to 0.35 within 2 seconds, exceeding the threshold of 0.25.

[0062] Mode switching and active intervention: The system detects P crisis ≥0.25, automatically switch to safety-first mode. The optimized solver, based on predictions from the simplified model, found that if the current pump rate is maintained, CHF... margin It will drop below 0.15 after 4 seconds, so a control command is calculated to increase the pump speed by 40%, even though x will be below 0.15 at this time. out It has not yet exceeded the standard.

[0063] Crisis resolution: After increasing the pump speed, the flow rate increased, CHF margin It started to recover, L b Stop decreasing and gradually recover, P crisis The value gradually dropped below 0.25. The system switched back to normal mode and smoothly adjusted the pump speed to the optimal value that balances temperature and energy consumption.

[0064] Through the above process, the system proactively intervenes before the outlet dryness exceeds the standard, successfully avoiding a potential membrane boiling crisis (burning dry), while achieving a balance between temperature control and energy consumption optimization.

[0065] like Figure 4 As shown in the figure, this application provides a power battery cooling control device, including: The signal acquisition module 401 is used to acquire measurement signals, including battery temperature, battery current, coolant pump speed, and cold plate outlet dryness. The first calculation module 402 is used to obtain the instantaneous heat generation rate of the battery based on the battery temperature, battery current and preset battery heat load estimation model; The second calculation module 403 is used to obtain the average dryness of the cold plate and the boiling front position based on the coolant pump speed, the dryness of the cold plate outlet and the pre-set two-phase flow dynamic model of the cold plate. The third calculation module 404 is used to calculate the critical heat flux density margin of the cold plate based on the instantaneous heat generation rate of the battery. The fourth calculation module 405 is used to calculate the boiling probability based on the cold plate outlet dryness, the average dryness of the cold plate, the boiling front position and the critical heat flux margin, using a pre-trained boiling prediction model. The control output module 406 is used to calculate the pump speed control command based on the average dryness of the cold plate, the position of the boiling front, and the boiling probability through a multi-objective model prediction controller; the pump speed control command is used to change the coolant flow rate to adjust the cooling speed of the power battery.

[0066] This application's solution achieves precise prediction and control of the two-phase flow boiling cooling process by constructing a closed-loop control architecture integrating perception, prediction, and decision-making. By reconstructing the unmeasurable boiling state inside the cold plate in real time, including average dryness fraction, boiling front position, and critical heat flux margin, the control system possesses a deep perception capability of the internal evolution of the boiling process. Based on this, by fusing the dynamic characteristics of the dryness fraction signal with the real-time estimated critical heat flux margin, a support vector machine classifier is used to achieve early probability prediction of film boiling risk, elevating the control strategy from traditional post-event response to pre-event warning. Finally, the optimization decision layer based on model predictive control can dynamically balance battery temperature control accuracy and pump power consumption while meeting absolute safety constraints, and automatically switch control modes according to risk probability, achieving an adaptive balance between safety priority and efficiency optimization. Therefore, this application fundamentally solves the technical defects of existing technologies that rely solely on delayed output signals and cannot perceive internal states or predict risks. It realizes real-time reconstruction of the internal state of the two-phase flow boiling cooling process, enabling advance quantitative prediction of risks and dynamic optimization control of the system. While ensuring the ultimate safety of the battery, it maximizes the system's energy efficiency and provides a breakthrough solution for thermal management in high-power battery fast charging scenarios.

[0067] This application also provides a computer-readable storage medium storing program information. After reading the program information, the computer executes the steps of the power battery cooling control method described in any of the above method embodiments.

[0068] This application also provides a computer program product, including a computer program / instructions, which, when executed by a processor, implement the steps of the power battery cooling control method described in any of the above method embodiments.

[0069] This application also provides an electronic device, such as... Figure 5 As shown, the electronic device includes at least one processor 51 and at least one memory 52. ​​The at least one memory 52 stores program information. After reading the program information, the at least one processor 51 executes the power battery cooling control method described in any of the above method embodiments. The device may further include an input device 53 and an output device 54. The processor 51, memory 52, input device 53, and output device 54 can be communicatively connected. The memory 52, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. The processor 51 executes various functional applications and data processing by running the non-volatile software programs, instructions, and modules stored in the memory 52, thereby implementing the power battery cooling control method provided in any of the above embodiments. The memory 52 may include a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created based on the use of the power battery cooling control method, etc. Furthermore, memory 52 may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 52 may optionally include memory remotely located relative to processor 51, and these remote memories may be connected via a network to the apparatus performing the power battery cooling control method. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof. Input device 53 may receive user clicks and generate signal inputs related to user settings and function control of the power battery cooling control method. Output device 54 may include a display device such as a display screen. When one or more modules are stored in memory 52 and are executed by one or more processors 51, the power battery cooling control method in any of the above method embodiments is performed.

[0070] This application also provides a vehicle that includes the power battery cooling control device described in the above embodiments, or a computer-readable storage medium, or a computer program product, or an electronic device.

[0071] As needed, the above technical solutions can be combined to achieve the best technical effect.

[0072] The above are merely the principles and preferred embodiments of this application. It should be noted that, for those skilled in the art, several other modifications can be made based on the principles of this application, and these modifications should also be considered within the scope of protection of this application.

Claims

1. A method for controlling the cooling of a power battery, characterized in that, include: Acquire measurement signals, including battery temperature, battery current, coolant pump speed, and cold plate outlet dryness; The instantaneous heat generation rate of the battery is obtained based on the battery temperature, battery current, and preset battery heat load estimation model. Based on the coolant pump speed, cold plate outlet dryness, and the two-phase flow dynamic model of the pre-installed cold plate, the average dryness of the cold plate and the boiling front position are obtained. Calculate the critical heat flux density margin of the cold plate based on the instantaneous heat generation rate of the battery; The boiling probability is calculated using a pre-trained boiling prediction model based on the cold plate outlet dryness, cold plate average dryness, boiling front position, and critical heat flux margin. Based on the average dryness of the cold plate, the position of the boiling front, and the boiling probability, a multi-objective model predicts the pump speed control command to calculate the pump speed control command. The pump speed control command is used to change the coolant flow rate to adjust the cooling rate of the power battery.

2. The power battery cooling control method according to claim 1, characterized in that, In the step of obtaining the instantaneous heat generation rate of the battery based on the battery temperature, battery current, and a preset battery heat load estimation model, the battery heat load estimation model is expressed as follows: ； in, Indicates the instantaneous heat generation rate of the battery. This indicates the battery current. This indicates the ohmic internal resistance of the battery. This indicates the battery temperature. It is the entropy heat coefficient.

3. The power battery cooling control method according to claim 2, characterized in that, The cold plate is divided into two regions along the refrigerant flow direction: a single-phase liquid region from the inlet to the boiling front, and a two-phase boiling region from the boiling front to the outlet; in the step of obtaining the average dryness of the cold plate and the boiling front position based on the coolant pump speed, the dryness of the cold plate outlet, and the pre-set two-phase flow dynamic model of the cold plate, the two-phase flow dynamic model of the cold plate is expressed as: ； ； in, This indicates the average dryness of the cold-rolled steel plate. The boiling front position; and These represent the densities of the gas and liquid phases, respectively. Indicates the latent heat of vaporization; Indicates the cross-sectional area of ​​the cold plate flow channel; L represents the length of the two-phase region, and L represents the total length of the cold plate. This represents the mass flow rate of the refrigerant, which is calculated from the pump characteristic curve based on the coolant pump speed. Indicates the dryness of the cold plate inlet; This represents the amount of heat absorbed by the battery in the two-phase region. Assuming the heat flux density is uniform, then... ; This represents the heat absorbed by the single-phase region. ; This represents the change in specific enthalpy of a single-phase liquid.

4. The power battery cooling control method according to claim 3, characterized in that, In the step of calculating the critical heat flux density margin of the cold plate based on the instantaneous heat generation rate of the battery, the critical heat flux density margin is calculated in the following manner: ； ； ； in, This indicates the critical heat flux density margin. Indicates surface tension. Indicates the dryness of the cold plate outlet. and Represents an empirical constant. This represents the current maximum heat flux density. A critical heat flux density margin greater than 0 indicates safety, while a margin less than 0 indicates that the material has burned out.

5. The power battery cooling control method according to claim 1, characterized in that, The step of calculating the boiling probability using a pre-trained boiling prediction model based on the cold plate outlet dryness, average cold plate dryness, boiling front position, and critical heat flux margin includes: The dryness of the cold plate outlet is sampled using a sliding window, and the time-domain statistical features and frequency-domain features of the signal within the window are extracted. The time-domain statistical features and the frequency-domain features are combined with the average dryness of the cold plate, the boiling front position, and the critical heat flux margin to form a multi-dimensional feature vector; The multidimensional feature vector is input into the pre-trained boiling prediction model, and the boiling prediction model outputs the boiling probability.

6. The power battery cooling control method according to claim 5, characterized in that: The boiling probability is the probability of film boiling occurring within a set time period in the future; the set time period is less than 10 seconds.

7. The power battery cooling control method according to claim 6, characterized in that, The boiling prediction model is a support vector machine classifier, which is represented as follows: ； in, The boiling probability is represented by H and B, which are pre-fitted constants. This represents the classification score of the support vector machine classifier.

8. The power battery cooling control method according to claim 1, characterized in that, In the step of calculating the pump speed control command using a multi-objective model predictive controller based on the average dryness of the cold plate, the position of the boiling front, and the boiling probability, the multi-objective model predictive controller includes: Control Model: ； ； in, This indicates the average dryness of the cold-rolled steel plate. This indicates the average wall temperature of the cold-rolled steel plate. Indicates the coolant pump speed. Indicates the instantaneous heat generation rate of the battery; a ij , b i , d i These represent the model coefficients, where i and j are both 1 or 2. and Indicates the period number; Objective function: ； in, Indicates the predicted future number The highest battery temperature of the step, Indicates the target battery temperature; α and β This represents the weighting coefficient.

9. The power battery cooling control method according to claim 8, characterized in that, The multi-objective model prediction controller also includes: Constraints: ,in ε This is a safety threshold; ； in, Indicates the predicted future number The critical heat flux margin of the step, and This indicates the lower and upper threshold values ​​for the coolant pump speed.

10. The power battery cooling control method according to any one of claims 1-9, characterized in that, The method further includes: The pump speed control command is represented as a control sequence. , with the first value As the optimal control value for the current cycle; When the boiling probability is less than the set threshold, the system enters the performance optimization mode. When the boiling probability is greater than or equal to the set threshold, the system enters a safety-first mode, and a penalty term is added to the objective function. , γ The penalty coefficient is increased; the safety threshold is raised, and the weighting is adjusted. α reduce.

11. A power battery cooling control device, characterized in that, include: The signal acquisition module is used to acquire measurement signals, including battery temperature, battery current, coolant pump speed, and cold plate outlet dryness. The first calculation module is used to obtain the instantaneous heat generation rate of the battery based on the battery temperature, battery current and the preset battery heat load estimation model. The second calculation module is used to obtain the average dryness of the cold plate and the boiling front position based on the coolant pump speed, the dryness of the cold plate outlet and the pre-set two-phase flow dynamic model of the cold plate. The first calculation module is used to calculate the critical heat flux density margin of the cold plate based on the instantaneous heat generation rate of the battery. The fourth calculation module is used to calculate the boiling probability based on the cold plate outlet dryness, the average dryness of the cold plate, the boiling front position, and the critical heat flux margin, using a pre-trained boiling prediction model. The control output module is used to predict and calculate the pump speed control command based on the average dryness of the cold plate, the position of the boiling front, and the boiling probability using a multi-objective model. The pump speed control command is used to change the coolant flow rate to adjust the cooling speed of the power battery.

12. A computer-readable storage medium, characterized in that, The storage medium stores program information, and after the computer reads the program information, it executes the steps of the power battery cooling control method according to any one of claims 1-10.

13. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the power battery cooling control method according to any one of claims 1-10.

14. An electronic device comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the power battery cooling control method according to any one of claims 1-10.

15. A vehicle, characterized in that, This includes the power battery cooling control device of claim 11, the computer-readable storage medium of claim 12, the computer program product of claim 13, or the electronic device of claim 14.

Images