Electric vehicle predictive thermal management control method, device, system, and storage medium

Abstract

The present invention discloses a predictive thermal management control method, device, system, and storage medium for electric vehicles, comprising: step S1, predicting operating condition data during vehicle driving using a DQN reinforcement learning method; wherein the operating condition data includes vehicle speed and battery current; step S2, based on the vehicle speed and battery current, using a global optimal algorithm through an upper-level control strategy to obtain an optimal battery temperature trajectory with the goal of minimizing energy loss, tracking the obtained optimal temperature trajectory through a lower-level control strategy, and controlling the electric vehicle thermal management system with the goal of optimizing the thermal management system's energy consumption. Using the technical solution of the present invention, the battery temperature is optimized and the thermal management system's energy consumption is minimized during driving, ultimately reducing system energy consumption and increasing the electric vehicle's driving range.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric vehicles, and in particular relates to a predictive thermal management control method and device, system, and storage medium for electric vehicles. Background Art

[0002] With the development of energy, electric vehicles are becoming more and more a trend in today's society. In order to alleviate the energy crisis and reduce carbon emissions, vigorously developing electric vehicles is a top priority. Among them, since the current thermal management system control strategies are mostly reactive control of current operating conditions, they lack the ability to predict future driving conditions and quickly respond to changes in vehicle speed to reduce the energy consumption of the thermal management system. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a predictive thermal management control method and device, system, and storage medium for electric vehicles.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] A predictive thermal management control method for an electric vehicle, comprising:

[0006] Step S1: predicting the operating condition data of the vehicle during driving by using the DQN reinforcement learning method; wherein the operating condition data includes: vehicle speed and battery current;

[0007] Step S2: Based on the vehicle speed and battery current, the upper-level control strategy uses a global optimal algorithm to obtain the optimal temperature trajectory of the battery with the goal of minimizing energy loss. The lower-level control strategy tracks the obtained optimal temperature trajectory, and an intelligent agent is used to coordinately control the components of the electric vehicle thermal management system with the goal of optimizing the energy consumption of the thermal management system.

[0008] Preferably, the components of the thermal management system include: a compressor, a condenser, an evaporator, a battery and a fan.

[0009] The present invention also provides a predictive thermal management control device for an electric vehicle, comprising:

[0010] The first processing module is used to predict the operating condition data of the vehicle during driving by using the DQN reinforcement learning method; wherein the operating condition data includes: vehicle speed and battery current;

[0011] The second processing module is used to obtain the optimal temperature trajectory of the battery based on the vehicle speed and battery current using the global optimal algorithm through the upper-level control strategy with the goal of minimizing energy loss. The optimal temperature trajectory is tracked through the lower-level control strategy, and an intelligent agent is used to collaboratively control the components of the electric vehicle thermal management system with the goal of optimizing the energy consumption of the thermal management system.

[0012] Preferably, the components of the thermal management system include: a compressor, a condenser, an evaporator, a battery and a fan.

[0013] The present invention also provides an electric vehicle predictive thermal management control system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes the electric vehicle predictive thermal management control method when executed by the processor.

[0014] The present invention also provides a storage medium having a computer program stored thereon, and the computer program executes the predictive thermal management control method for an electric vehicle when running.

[0015] The present invention adopts a hierarchical control strategy based on driving conditions, which makes the battery temperature reach the optimal state and the energy consumption of the thermal management system reach the minimum during driving, ultimately reducing the system energy consumption and increasing the driving range of the electric vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0017] Figure 1 This is a flow chart of a predictive thermal management control method for electric vehicles according to an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] Example 1:

[0021] like Figure 1 As shown, an embodiment of the present invention provides a predictive thermal management control method for an electric vehicle, comprising:

[0022] Step S1: predicting the operating condition data of the vehicle during driving by using the DQN reinforcement learning method; wherein the operating condition data includes: vehicle speed and battery current;

[0023] Step S2: Based on the vehicle speed and battery current, the upper-level control strategy uses the global optimal algorithm to obtain the optimal temperature trajectory of the battery with the goal of minimizing energy loss. The optimal temperature trajectory is tracked by the lower-level control strategy, and the components of the electric vehicle thermal management system are collaboratively controlled by an intelligent agent with the goal of optimizing the energy consumption of the thermal management system; wherein, the components of the thermal management system include: compressor, condenser, evaporator, battery and fan. In step S2, the system current is calculated based on the vehicle speed information, and the internal heat generation power of the battery is further derived, and then the dynamic change trajectory of the battery temperature is derived from the thermal balance equation. The vehicle speed change has a direct coupling relationship with the battery temperature, and the vehicle speed-temperature mapping function can be constructed under different working conditions to provide predictive support for subsequent control strategies. The temperature model supports the optimal trajectory planning of the upper-level control strategy, and realizes the collaborative adjustment of the thermal management system components through the lower-level intelligent agent to achieve the optimal comprehensive energy consumption goal. The specific steps are as follows: the driving force of the vehicle during driving is determined by the vehicle dynamics model, and the battery current ,in, For driving power, is the battery voltage, is the vehicle speed, To drive system efficiency, = ,in 、 、 ,According to the Joule heating effect, the heat generated by the battery is caused by the internal resistance: ,in: Heat production per unit time, is the internal resistance of the battery. Then, according to the law of conservation of energy, the temperature rise is determined by the difference between heat generation and heat dissipation: , the integral gives the temperature change over time: = ,in is the battery temperature, Heat dissipation power for cooling system, For battery quality, is the battery specific heat capacity, given by This relationship gives the battery temperature change ， In this way, the upper-level control strategy can use the global optimal algorithm to obtain the optimal temperature trajectory of the battery with the goal of minimizing energy loss. The optimal temperature trajectory can be tracked by the lower-level control strategy, and the intelligent agent can be used to coordinately control the components of the electric vehicle thermal management system with the goal of optimizing the energy consumption of the thermal management system.

[0024] As an implementation of an embodiment of the present invention, in step S1, five elements of reinforcement learning are determined:

[0025] (1) State space : The current state of the environment in which the agent is located. For vehicle driving condition prediction, the state can include vehicle speed , acceleration , throttle opening , steering angle , ambient temperature , battery temperature ,Right now , these state variables form the input of the reinforcement learning model, and the agent needs to predict future driving conditions based on this information.

[0026] (2) Action Space : Actions that the agent can choose. In driving condition prediction, actions might include adjusting the throttle position, braking pressure, and steering angle. The agent selects the optimal action based on the current state, making the vehicle's driving prediction more accurate and ensuring efficient operation of the powertrain.

[0027] (3) Reward Function : The feedback obtained after the agent performs a certain action, the error between the predicted speed and the actual speed is selected as the reward function ,in, represents the predicted acceleration, Represents the weight of controlling acceleration changes. The reward function is used to measure the quality of the current action. This reward function ensures that the agent can minimize the error when predicting driving conditions while optimizing energy consumption and driving stability.

[0028] (4) Strategy :The agent is based on the state Select Action The probability distribution of , where estimated by the Deep Q-Network (DQN), is the temperature parameter.

[0029] (5) Value function :Indicates that the state Take action The expected cumulative reward after .

[0030] DQN is a reinforcement learning method based on Q-learning, which combines deep neural networks (DNN) to approximate the Q-value function. , which solves the difficulties of traditional Q-learning in high-dimensional state space. The core idea of Q-learning is to update the state-action value function , each action chosen by the agent will have an impact on the future cumulative reward. The goal is to maximize the accumulated long-term reward.

[0031] The update formula of the Q value function is:

[0032] ;

[0033] in, and respectively represent time The state and action of For the moment Execute an action Rewards received after is the discount factor, controlling the impact of future rewards; is the learning rate, which controls the step size of Q value update.

[0034] DQN uses deep neural networks To approximate the Q value function, where When updating the Q value, DQN uses experience replay and the target network to improve the stability of training.

[0035] Experience replay: Each interaction of the agent generates an experience tuple ,These tuples are stored in the experience pool, and a batch of samples are randomly selected to train the neural network to reduce the correlation between the data.

[0036] Target network: In order to improve the stability of training, DQN uses a target network , the parameters of this network Regularly from online networks copy.

[0037] The goal of updating the Q value is:

[0038] As an implementation of an embodiment of the present invention, in step S1, predicting the operating condition data of the vehicle during driving by using a deep Q-learning (DQN) method includes:

[0039] Initialization phase:

[0040] S11. Initialize the Q value network and target network , and set the initial parameters and ;

[0041] S12. Initialize the experience replay pool ;

[0042] S13. Set learning rate , discount factor , and the interval between target network updates;

[0043] Training process stages:

[0044] S14. Set initial state ;

[0045] S15. Select action: Select an action based on the current strategy , using the ϵ-greedy policy:

[0046] ;

[0047] in, is the exploration rate, which is used to balance exploration and exploitation;

[0048] S16. Execute action: Execute action , and get rewards based on feedback from the environment and the next state ;

[0049] S17, storage experience: experience tuple Store in experience replay pool ;

[0050] S18, Update Q value: Randomly sample a batch of experience from the experience pool , calculate the target value , , by minimizing the loss function To update the parameters of the Q value network , update the target network every fixed number of steps. ,

[0051] S19. Update strategy: As training progresses, update the strategy and continue execution.

[0052] S20, Convergence Verification: Finally, convergence and testing are performed. The training process will continue until the model converges, that is, the Q value is stable and the desired control strategy is achieved. In the testing phase, the intelligent agent performs actions according to the learned strategy, evaluates its performance in the actual environment, adjusts the strategy based on the training results, and then trains until the results converge.

[0053] As an implementation of the embodiment of the present invention, in step S2, the battery temperature is selected. As system state variables , select the battery circuit cooling capacity As system control quantity

[0054] Establishing the equation of state for thermal management systems

[0055] ;

[0056] in, is the battery power demand at time t, is the cooling work of the compressor at time t, is the instantaneous rate of change of battery temperature.

[0057] With the goal of global temperature control and energy consumption minimization, the corresponding objective function is established as

[0058] in, ,in, Heat dissipation for the battery, is the state cost coefficient in the dynamic algorithm, is the cooling system efficiency matrix, which is used to calculate the system cooling efficiency, 、 To control the cost coefficient in the dynamic algorithm, the values of each variable are restricted. The following constraints need to be met during the entire algorithm iteration process: .

[0059] Based on a dynamic algorithm, the hierarchical iterative algorithm decomposes complex problems into multiple levels, each corresponding to a subproblem, ensuring clear dependencies between subproblems. During the solution process, after initialization, the hierarchical iterations are repeated. During the upper-level solution, the current lower-level solution is fixed, the upper-level problem is solved, and the upper-level variables are updated. During the lower-level solution, the updated results of the upper-level solution are used to optimize the lower-level problem and solve the lower-level variables. When convergence conditions are met, the iteration terminates; otherwise, the optimization returns to the upper level and continues. It is The objective function of the stage is expressed in the state The specific goal is to find the optimal solution of the system by minimizing this goal. It is The control strategy of the stage is expressed in the state The optimal control quantity under stage, state The optimal control quantity under By minimizing the cost function The optimal control quantity obtained :

[0060] ;

[0061] in, Represents the number of optional control strategies.

[0062] Through the recursive relationship The optimal cost function of the calculation stage is: Based on the status of the previous stage and control The calculated cost function is The current stage The cost function represents the next state In order to ensure smooth transitions between states and optimized control strategies at different stages, it is necessary to further introduce constraints on states and controls, using the following formula:

[0063] ;

[0064] Among them, in the state Downstream Execution Control After that, get the status Certain constraints must be met to ensure that state changes are consistent with system dynamics.

[0065] By optimizing the control strategy to maintain the optimal operating temperature of the battery, in the state transformation and control strategy update, the following formula:

[0066] ;

[0067] Through the proportional gain and the power of the temperature error Control and maintain the battery temperature in the optimal temperature range.

[0068] Furthermore, a multi-agent system is used to manage the air conditioning system in electric vehicles to ensure that battery thermal management is maintained at optimal energy efficiency.

[0069] The compressor agent uses deep reinforcement learning (DDPG), which is suitable for nonlinear systems and obtains the required cooling capacity according to the upper-level control algorithm. , input the compressor speed and the current battery thermal management system power demand constraint, output the compressor power It also calculates the pressure ratio of the evaporator and condenser, and interacts with the electronic expansion valve, condenser fan, and battery agent. The condenser agent uses deep Q learning (DQN) to better apply to the current mainstream condenser fan gear control. The agent inputs the target condensing temperature and accepts the high-temperature and high-pressure gas output from the compressor. It uses the condenser fan speed and condensing temperature as outputs, and coordinates with the compressor agent to adjust the heat exchange efficiency. The evaporator agent uses the PPO algorithm to cope with complex environments and handle the continuous control of the expansion valve opening. It uses the passenger compartment temperature, humidity, and evaporator target temperature as input. Output the opening of the electronic expansion valve and calculate the refrigerant mass flow rate, control the refrigerant flow, improve the evaporation efficiency and coordinate with the fan's intelligent body to adjust the air temperature. The fan intelligent body adopts the TD3 algorithm, which is suitable for complex environments and can cope with the continuous speed adjustment of the fan. It takes the supply air temperature and the in-vehicle temperature sensor data as input, outputs the supply air volume, and interacts with the evaporator, etc. Increasing the wind speed can reduce the evaporator temperature, and reducing the wind speed can improve passenger comfort. The battery intelligent body adopts the SAC reinforcement learning method, which is suitable for complex battery cooling systems, which is conducive to optimizing the coolant flow control and using the current battery temperature , Maximum safe temperature of battery , ambient temperature As input, adjust the coolant flow and start battery cooling. When the battery temperature When the vehicle is in a cool state, the air conditioning system is started for cooling and the compressor power is adjusted to prevent the battery from overheating.

[0070] Agent training process: 1. The compressor agent uses the deep deterministic policy gradient algorithm (DDPG) to dynamically adjust the compressor power, reduce energy consumption, and ensure the stability of battery temperature. , target temperature and ambient temperature As the input state, the power range of the compressor is 0 to 5000W as the action decision, and the reward function is designed to be The training is conducted in a pre-built training environment. The initial state is set to the outdoor temperature and the compressor power is randomly initialized. The agent learns the optimal power regulation strategy through trial and error. The training is conducted until the reward converges, and the performance of the model is verified under different temperature conditions. 2. The condenser agent aims to ensure that the condenser is kept at the optimal temperature by adjusting the condenser fan speed. , target temperature and ambient temperature Taking the fan speed as the action space for the state input, the reward function is designed as , using the deep Q network algorithm, first the initial state is set to the initial temperature of the condenser, the fan is randomly started, and the agent learns how to adjust the fan speed to optimize the condensing efficiency, training until the reward converges, and evaluating the performance under different external temperatures. 3. The evaporator agent aims to adjust the refrigerant flow, control the temperature inside the car, and maintain passenger comfort. , Humidity inside the car Taking the refrigerant flow rate as the action space for the state input, the reward function is designed as in The intelligent agent is trained to dynamically adjust the refrigerant flow rate for the refrigerant mass flow rate, find the best balance between energy saving and comfort, train to minimize the temperature error, and evaluate the energy saving effect. 4. The battery intelligent agent controls the coolant flow rate to ensure that the battery temperature is stable at 15°C ~ 40°C. , target temperature , ambient temperature is the state input, with coolant flow Design a reward function for the action space , using the SAC algorithm and the battery thermal model to simulate the battery temperature changes under different driving conditions. 5. The fan agent uses the current temperature of the passenger compartment to , Current humidity in the passenger compartment , target temperature , Current fan speed Fan energy consumption Design a reward function for the action space , while ensuring passenger comfort, minimizing fan energy consumption

[0071] Train an agent to control coolant flow and optimize battery temperature.

[0072] In order to achieve accurate tracking of the battery temperature trajectory, a deep learning method is used to optimize the control strategy so that the battery temperature is close to the target temperature trajectory and the temperature error and energy consumption are minimized. First, the system modeling and state space are defined, and a battery system model is established, especially the dynamic characteristics of battery temperature management, including the temperature error in the current system state ( ) and the cooling capacity of the battery circuit as the control quantity Used to regulate battery temperature and to capture dynamic historical temperature information of the system 、 And the previous cooling capacity, etc., to form time series data. Through the above state variables, define the current system state , using this information as the input of the deep learning model to help the model learn the dynamic changes of battery temperature. Then define the action space. The agent needs to make a decision based on the current state and select an action. In this case, we choose the change in the cooling capacity of the battery circuit as the action space. ,in, is an action that represents a change in cooling capacity (for example, increasing or decreasing the cooling capacity of the battery circuit). The action space is a continuous space, so the policy network of deep reinforcement learning can be used to learn the selection of actions. The reward function is then designed. The reward function is the key in reinforcement learning, which determines the learning goal of the intelligent agent. In this problem, our goal is to minimize the temperature error and keep the energy consumption of the thermal management system to the lowest. Therefore, we need to design a reward function that combines temperature error and energy consumption. Battery cooling capacity A certain amount of energy will be consumed, which can be expressed as the energy consumption of the system: in is a function of energy consumption, usually related to cooling capacity Proportional to the designed reward function It can be expressed as a trade-off between the penalty for temperature error and energy consumption: ,in is the energy consumption penalty factor, is the energy consumption of the thermal management system at the current moment, is the penalty factor for temperature error, controlling the importance of temperature error in rewards. After that, the deep learning model is selected and trained, and deep reinforcement learning (DRL) is used to learn the optimal control strategy. The agent continuously updates the policy network by interacting with the environment to minimize temperature error and energy consumption. During the training process, due to the movement of space Non-discrete is continuous, so using the deep Q network (DQN) of Q-learning, the agent selects the optimal action by maximizing future rewards. Based on the current state Execute an action , get a new state from the environment and rewards This process will continue and accumulate experience. The goal of the reinforcement learning algorithm to update the Q value is: Among them, the target value , is the Q-value network, For the target network, and is the initial parameter learning rate , The training process updates the neural network parameters through back propagation, thereby continuously improving the decision-making ability of the intelligent agent. The intelligent agent learns how to adjust the cooling capacity according to the current temperature error and system status, thereby minimizing the temperature error and reducing energy consumption. The training process updates the neural network parameters through back propagation, thereby continuously improving the decision-making ability of the intelligent agent. After the training is completed, the intelligent agent will be able to adjust the battery circuit cooling capacity in real time according to the current temperature error and historical status. For each time step , the agent is based on the current state Predict the control strategy for the next moment , and adjust system parameters (such as cooling capacity) to make the temperature approach the target trajectory. Specifically, the intelligent experience will and current temperature Calculate the temperature error Then adjust the strategy based on the training , and continuously corrected and optimized in subsequent steps. The training goal is to minimize the total temperature error throughout the entire process. To achieve this, the temperature error can be weighted averaged or the error at multiple moments can be optimized through cumulative rewards to ensure that the temperature trajectory is always within the optimal range. The following optimization objectives are used during training:

[0073] in, is the total training time, and the optimization goal is to minimize the total temperature error in the entire process. After training, the intelligent agent will be able to control the cooling capacity of the battery circuit in real time to make the battery temperature as close to the target trajectory as possible. When the system is deployed, the model should be strictly verified and evaluated to ensure that it can show good stability and adaptability under different loads and different environmental conditions. Evaluation indicators may include: (1) Temperature error: whether the temperature error is kept within a reasonable range during the entire training process. (2) System energy consumption: whether the system energy consumption is lower than the preset threshold and the use of cooling capacity is optimized as much as possible. (3) System response time: the speed and accuracy of temperature adjustment to ensure that the system can quickly respond to changes in battery temperature. Optimizing the battery temperature control strategy through deep reinforcement learning can not only achieve accurate temperature trajectory tracking, but also effectively reduce energy consumption and improve the stability and adaptability of the system. The ultimate goal is to enable the battery temperature management system to adaptively adjust in various actual environments through continuous learning of the intelligent agent to ensure that the battery operates in the optimal temperature range, thereby extending battery life and improving system efficiency.

[0074] Example 2:

[0075] An embodiment of the present invention further provides a predictive thermal management control device for an electric vehicle, comprising:

[0076] The first processing module is used to predict the operating condition data of the vehicle during driving by using the DQN reinforcement learning method; wherein the operating condition data includes: vehicle speed and battery current;

[0077] The second processing module is used to obtain the optimal temperature trajectory of the battery based on the vehicle speed and battery current using the global optimal algorithm through the upper-level control strategy with the goal of minimizing energy loss. The optimal temperature trajectory is tracked through the lower-level control strategy, and an intelligent agent is used to collaboratively control the components of the electric vehicle thermal management system with the goal of optimizing the energy consumption of the thermal management system.

[0078] As an implementation manner of an embodiment of the present invention, components of a thermal management system include: a compressor, a condenser, an evaporator, a battery, and a fan.

[0079] Example 3:

[0080] The present invention also provides an electric vehicle predictive thermal management control system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes the electric vehicle predictive thermal management control method when executed by the processor.

[0081] Example 4:

[0082] The present invention also provides a storage medium having a computer program stored thereon, and the computer program executes the predictive thermal management control method for an electric vehicle when running.

[0083] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A predictive thermal management control method for electric vehicles, characterized in that: include: Step S1: predicting the operating condition data of the vehicle during driving by using the DQN reinforcement learning method; wherein the operating condition data includes: vehicle speed and battery current; Step S2: Based on the vehicle speed and battery current, the upper-level control strategy uses a global optimal algorithm to obtain the optimal temperature trajectory of the battery with the goal of minimizing energy loss. The lower-level control strategy tracks the obtained optimal temperature trajectory, and an intelligent agent is used to coordinately control the components of the electric vehicle thermal management system with the goal of optimizing the energy consumption of the thermal management system. In step S1, the operating condition data of the vehicle during driving is predicted using a deep Q learning method, including: Initialization phase: S11. Initialize the Q value network and target network , and set the initial parameters and ; S12. Initialize the experience replay pool ; S13. Set learning rate , discount factor , and the interval between target network updates; Training process stages: S14. Set initial state ; S15. Select action: Select an action based on the current strategy ,use -Greedy strategy: ； in, is the exploration rate, which is used to balance exploration and exploitation; S16. Execute action: Execute action , and get rewards based on feedback from the environment and the next state ; S17, storage experience: experience tuple Store in experience replay pool ; S18, Update Q value: Randomly sample a batch of experience from the experience pool , calculate the target value , , by minimizing the loss function To update the parameters of the Q value network , update the target network every fixed number of steps, ; S19, Update strategy: As training progresses, update the strategy and continue execution; S20, Convergence Verification: Finally, convergence and testing are performed. The training process will continue until the model converges, that is, the Q value is stable and the desired control strategy is achieved. In the testing phase, the intelligent agent performs actions according to the learned strategy, evaluates its performance in the actual environment, adjusts the strategy based on the training results, and then trains until the results converge.

2. The predictive thermal management control method for electric vehicles according to claim 1, wherein: Components of the thermal management system include: compressor, condenser, evaporator, battery and fan.

3. An electric vehicle predictive thermal management control device that implements the electric vehicle predictive thermal management control method according to claim 1, characterized in that: include: The first processing module is used to predict the operating condition data of the vehicle during driving by using the DQN reinforcement learning method; wherein the operating condition data includes: vehicle speed and battery current; The second processing module is used to obtain the optimal temperature trajectory of the battery based on the vehicle speed and battery current using the global optimal algorithm through the upper-level control strategy with the goal of minimizing energy loss. The optimal temperature trajectory is tracked through the lower-level control strategy, and an intelligent agent is used to collaboratively control the components of the electric vehicle thermal management system with the goal of optimizing the energy consumption of the thermal management system.

4. The predictive thermal management control device for an electric vehicle according to claim 3, wherein: Components of the thermal management system include: compressor, condenser, evaporator, battery and fan.

5. A predictive thermal management control system for electric vehicles, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the electric vehicle predictive thermal management control method according to any one of claims 1 to 2 is executed.

6. A storage medium, characterized in that The storage medium stores a computer program, which, when running, executes the predictive thermal management control method for an electric vehicle as described in any one of claims 1 to 2.

Images