A control method, device and thermal management system for a vehicle thermal management system

Abstract

The present invention relates to the technical field of thermal management systems, and in particular, to a control method, device and thermal management system for a vehicle thermal management system. The method includes: pre-training a neural network model according to the bench test data of the electric drive thermal management system; inputting the parameters of the electric drive thermal management system into the neural network model to predict the outlet water temperature of the electric drive thermal management system; the parameters of the electric drive thermal management system include the water pump speed; using the neural network model as the prediction model of model predictive control to iterate the water pump speed to obtain the optimal water pump speed; and controlling the water pump with the optimal water pump speed. The present invention can perform refined control on the water pump speed.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal management systems, and more particularly, to a control method, a device, and a thermal management system for a vehicle thermal management system. Background Art

[0002] Taking the electric drive thermal management system of new energy vehicles as an example, this system is usually composed of a motor controller, a motor, a radiator, a water pump, and possibly a DC converter and a charger connected in series. The heat generated by the motor controller, the motor, and other electrical components such as the DC converter and the charger during operation is carried by the coolant in the thermal management system through the radiator driven by the water pump to exchange heat with the outside air, ensuring that the electrical components operate at an appropriate and safe temperature. Among them, the water pump is an energy-consuming component, and its speed control is usually a stepped control determined according to the water temperature of the electric drive circuit or the motor temperature.

[0003] Stepped control is difficult to accurately control the speed of the water pump, making it difficult to achieve a balance between the motor cooling effect and the water pump energy consumption.

[0004] In view of this, the present invention is specifically proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide a control method, a device, and a thermal management system for a vehicle thermal management system to finely control the speed of the water pump.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a control method for a vehicle thermal management system, including:

[0008] Pre-training a neural network model according to the bench test data of the electric drive thermal management system;

[0009] Inputting the parameters of the electric drive thermal management system into the neural network model to predict the outlet water temperature of the electric drive thermal management system; the parameters of the electric drive thermal management system include the water pump speed;

[0010] Using the neural network model as a prediction model for model predictive control to iterate the water pump speed to obtain the optimal water pump speed;

[0011] Controlling the water pump with the optimal water pump speed.

[0012] In a second aspect, the present invention provides a control device for a vehicle thermal management system, including:

[0013] A training module for pre-training a neural network model according to the bench test data of the electric drive thermal management system;

[0014] A prediction module, configured to input the parameters of the electric drive thermal management system into the neural network model to predict the outlet water temperature of the electric drive thermal management system; the parameters of the electric drive thermal management system include the water pump speed;

[0015] An iteration module, configured to use the neural network model as the prediction model of model predictive control to iterate the water pump speed to obtain the optimal water pump speed;

[0016] A control module, configured to control the water pump by using the optimal water pump speed.

[0017] In a third aspect, the present invention provides a thermal management system, characterized by comprising: an electronic device, a water pump, a radiator, a motor, a motor controller, and a DC converter;

[0018] Wherein, the electronic device includes: at least one processor, and a memory communicatively connected to at least one of the processors;

[0019] The memory stores instructions executable by at least one of the processors, and the instructions are executed by at least one of the processors to enable at least one of the processors to execute the vehicle thermal management system control method.

[0020] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0021] The present invention innovatively replaces the state space model with a neural network model to realize the physical simulation of the thermal management system. Based on this, the neural network model is used as the prediction model of model predictive control to iterate the water pump speed to obtain the optimal water pump speed, realizing the refined control of the water pump speed, and at the same time achieving the balance between the motor cooling effect and the water pump energy consumption. There is no solution in the prior art to control the water pump speed with this index and prediction model algorithm. Description of the Drawings

[0022] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0023] Figure 1 is a schematic flowchart of a vehicle thermal management system control method provided by an embodiment of the present invention;

[0024] Figure 2 is a schematic structural diagram of a bench test provided by an embodiment of the present invention;

[0025] Figure 3 It is a schematic diagram of the optimized process provided by an embodiment of the present invention;

[0026] Figure 4 It is a schematic structural diagram of a vehicle thermal management system control device provided by an embodiment of the present invention;

[0027] Figure 5 It is a schematic structural diagram of an electronic device provided by the present invention. Detailed implementation manners

[0028] The following describes exemplary embodiments of the present invention with reference to the accompanying drawings. Various details of the embodiments of the present invention are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted below.

[0029] Figure 1 It is a flowchart of a vehicle thermal management system control method provided by this embodiment. This embodiment is applicable to the situation of controlling the speed of a water pump in an electric drive thermal management system of a new energy vehicle. This method can be executed by an electronic device. Refer to Figure 1 In this embodiment, the method provided includes the following operations:

[0030] S110. Pre-train the neural network model according to the bench test data of the electric drive thermal management system.

[0031] The neural network model in this embodiment is preferably a Long Short-Term Memory (LSTM) model, which is specifically designed to solve the problems of gradient disappearance and gradient explosion that occur in traditional recurrent neural networks when processing long sequence data. LSTM can effectively capture long-term dependencies in time series data by introducing memory units and gating mechanisms.

[0032] In this embodiment, first, obtain the bench test data for training the neural network model.

[0033] Figure 2It is a schematic structural diagram of a bench test provided by an embodiment of the present application, including an environmental control system, a radiator, a water pump, a direct current (DC-DC) converter, a motor controller, a motor, and a dynamometer. Optionally, based on actual road condition tests, the electric drive thermal management system is mounted on the bench for testing, and bench test data is collected. Specifically, the bench dynamometer (located on both sides of the motor) provides road load simulation for the motor, and the environmental control system of the bench provides air intake volume and temperature control for the radiator. The bench can collect data such as the rotational speed and torque output by the electric drive system (i.e., the motor in this embodiment), the input voltage and current, the inlet water temperature and outlet water temperature of the electric drive thermal management system, the coolant flow rate, the water pump rotational speed, and the motor body temperature. In this embodiment, the bench test data includes the heat generation of the electric drive system, the water pump rotational speed, the inlet water temperature of the electric drive thermal management system, the motor body temperature, and the outlet water temperature of the electric drive thermal management system. Among them, the heat generation of the electric drive system can be obtained by subtracting the output mechanical power from the input electric power of the electric drive system.

[0034] During the test process, transient condition tests are generally included. Since the vehicle speed in these conditions always changes with time. Therefore, the input electric power and output mechanical power of the electric drive system, the water pump rotational speed, the flow rate and water temperature of the coolant, etc. also always change. After discretizing these bench test data at equal time intervals (such as 1 s), they can be used as learning samples for neural network model training. Specifically, taking the heat generation of the electric drive system, the inlet water temperature, the motor body temperature, and the water pump rotational speed as inputs, for example, the input is a time series of T1 to T10, and taking the outlet water temperature as the output label, the neural network model is pre-trained. The outlet water temperature will be used as the reference temperature for the water pump speed control strategy.

[0035] S120. Input the parameters of the electric drive thermal management system into the neural network model to predict the outlet water temperature of the electric drive thermal management system.

[0036] After the neural network model is trained, it is necessary to input the parameters of the actual electric drive thermal management system into the neural network model to predict the outlet water temperature at the next moment. Among them, the parameters of the electric drive thermal management system include the heat generation of the electric drive system, the inlet water temperature, the motor body temperature, and the water pump rotational speed collected at the current moment. The model output is the outlet water temperature at the next moment, that is, the predicted outlet water temperature.

[0037] S130. Use the neural network model as the prediction model of model predictive control to iterate the water pump rotational speed to obtain the optimal water pump rotational speed.

[0038] The innovation of the present invention lies in using a neural network model as the prediction model for the model predictive control (MPC) of an electronic water pump. In the prior art, the MPC model relies on a prediction model established based on a state space model. The implementation of MPC model predictive control is divided into three main links: First, a prediction model is constructed to predict the future behavior of the system; second, rolling optimization, that is, optimizing the problem within a finite time domain to obtain a control sequence; finally, feedback correction, by collecting system information in real time, correcting the prediction model, and continuously optimizing. The establishment of the state space model often used in the MPC control strategy requires detailed thermal management loop parameters and a complete physical model, and this model is prone to errors due to component aging, changes in the performance of electrical components, etc. in the later stage, affecting the MPC control strategy, and it is difficult to accurately predict the motor temperature using traditional state space equations. However, the neural network has strong non-linear fitting ability. Using an LSTM neural network as the prediction model on the basis of traditional MPC can avoid the above problems.

[0039] Optionally, the particle swarm optimization (PSO) algorithm is used to optimize the water pump speed to achieve a better thermal management effect. The PSO algorithm can search in the entire search space and is not easily trapped in local optimal solutions, which is very important for complex non-linear and non-convex optimization problems in MPC and is more likely to find the globally optimal control strategy. Compared with some traditional optimization algorithms, the principle and implementation of the PSO algorithm are relatively simple and do not require complex mathematical derivations and calculations, such as not requiring the calculation of the derivative of the objective function, etc., reducing the difficulty of algorithm development and application. Specifically, the water pump speed represented by each particle is iterated multiple times, and the optimized outlet water temperature is obtained through the neural network model after each iteration; the temperature reduction amplitude of the motor is obtained based on the outlet water temperature and the motor body temperature; the optimal water pump speed is selected according to the water pump speed represented by each particle and the temperature reduction amplitude of the motor.

[0040] The following is a specific implementation manner. Refer to Figure 3 , and the application of the PSO algorithm in the water pump speed control scenario will be described in detail.

[0041] Step 1: Initialize parameters. The prediction time domain and the control time domain are consistent with the dimension D of the particle. In this embodiment, it is set to 10, the maximum number of iterations is 50, the number of particles N is 75, and parameters such as position and velocity in the PSO algorithm.

[0042] The first dimension in the particle is the current temperature of the motor body. The second dimension in the particle is the current inlet water temperature. The third dimension in the particle is the heat generation of the current electric drive system. The fourth dimension in the particle is the current outlet water temperature. The fifth dimension in the particle is the motor body temperature at the previous time step. The sixth dimension in the particle is the inlet water temperature at the previous time step. The seventh dimension in the particle is the heat generation of the electric drive system at the previous time step. The eighth dimension in the particle is the outlet water temperature at the previous time step. The ninth dimension in the particle is the current pump speed. The tenth dimension in the particle is the optimized pump speed.

[0043] Step 2: Update the positions of the particles. First, update each component of the velocity vector of each particle according to the inertia weight obtained by each particle, and then update each component of the position vector according to the updated result.

[0044] Step 3: Predict the pump speed. Consider each updated particle in Step 3 as a potential feasible control sequence, and use the trained LSTM prediction model to predict the motor temperature after control, and output the corresponding pump speed, etc. for each particle.

[0045] Among them, the updated particles include the heat generation of the electric drive system, the pump speed, the inlet water temperature of the electric drive thermal management system, and the motor body temperature. Among them, the pump speed is updated with each iteration, but the heat generation of the electric drive system, the inlet water temperature, and the motor body temperature use the current parameters of the electric drive thermal management system and the historical data of the previous time step in previous iterations.

[0046] Step 4: Calculate the fitness. Calculate the fitness value corresponding to each group of particles according to the designed cost function, that is, evaluate the control effect of the control sequence represented by each particle within the prediction time domain according to the cost function.

[0047] Optionally, use the following as the cost function J:

[0048] ;

[0049] Among them, and are weights, which can be set according to actual needs. is the pump speed, that is, the pump speed represented by the particle after iteration. In this embodiment, 75 pump speeds can be obtained through iteration. is the motor body temperature, is the outlet water temperature.

[0050] If the cost function of a particle after iteration reaches the minimum or the number of iterations reaches the threshold, then use the pump speed represented by the particle as the optimal pump speed.

[0051] Step 5: Determine whether the iteration condition is met. If not, return to Step 2 and continue the iteration until the maximum number of iterations is reached. Denote the vector represented by the best historical position of the particle swarm (i.e., the position with the minimum cost function) as the control vector.

[0052] Step 6: Motor speed control. Take the last position in the obtained control vector (representing the optimal water pump speed) as the control value and input it into the actual motor control to complete the control of the current control cycle.

[0053] S140: Control the water pump with the optimal water pump speed.

[0054] Through experiments, it is proved that the simulation working condition is the NEDC (New European Driving Cycle) working condition, the ambient temperature is 30°C, and the control target is that the outlet water temperature is 32°C. The results show that the control error of the outlet water temperature can be kept within 2%, and an energy-saving effect of 3% can be achieved.

[0055] The present invention innovatively replaces the state space model with a neural network model to realize the physical simulation of the thermal management system. Based on this, the neural network model is used as the prediction model of model predictive control to iterate the water pump speed to obtain the optimal water pump speed, realizing the refined control of the water pump speed, and at the same time achieving the balance between the motor cooling effect and the water pump energy consumption. There is no existing technology that controls the water pump speed with this index and prediction model algorithm.

[0056] Replace the state space model with a neural network model as the prediction model of the electronic water pump MPC control strategy. The advantages of this method include:

[0057] 1. The real road conditions are originally the commonly used test conditions in the development and test process. Only need to organize the data obtained from the test without adding additional test links.

[0058] 2. Omit the complex mathematical modeling of establishing the state space model, and do not require the detailed design parameters of the electric drive system. 3. This neural network can be updated and learned in real time. When the radiator state and the performance of electrical components change, the neural network can be updated to ensure the accuracy of the MPC control strategy prediction model.

[0059] 3. Model predictive control (MPC) can control the water pump speed more refinedly. When controlling the water temperature of the thermal management loop and the motor temperature not to exceed the temperature limit, it can save more energy of the thermal management system.

[0060] Figure 4It is a schematic structural diagram of a control device for a vehicle thermal management system provided by an embodiment of the present invention, including a training module 201, a prediction module 202, an iteration module 203, and a control module 204.

[0061] The training module 201 is configured to pre-train a neural network model according to the bench test data of the electric drive thermal management system.

[0062] The prediction module 202 is configured to input the parameters of the electric drive thermal management system into the neural network model to predict the outlet water temperature of the electric drive thermal management system; the parameters of the electric drive thermal management system include the water pump speed.

[0063] The iteration module 203 is configured to use the neural network model as a prediction model for model predictive control to iterate the water pump speed to obtain the optimal water pump speed.

[0064] The control module 204 is configured to control the water pump by using the optimal water pump speed.

[0065] Optionally, the training module 201 is configured to, based on actual road condition tests, mount the electric drive thermal management system on a bench for testing and collect bench test data; wherein, the bench test data includes the heat generation of the electric drive system, the water pump speed, the inlet water temperature of the electric drive thermal management system, the motor body temperature, and the outlet water temperature of the electric drive thermal management system; the bench dynamometer provides a road load simulation for the motor, and the environmental control system of the bench provides the air intake volume and temperature control for the radiator; using the heat generation of the electric drive system, the inlet water temperature, the motor body temperature, and the water pump speed as inputs and the outlet water temperature as the output label, pre-train the neural network model.

[0066] Optionally, the iteration module 203 is configured to use the particle swarm optimization algorithm to iterate the water pump speed represented by each particle multiple times, and obtain the optimized outlet water temperature through the neural network model after each iteration; obtain the temperature reduction amplitude of the motor according to the outlet water temperature and the motor body temperature; select the optimal water pump speed according to the water pump speed represented by each particle and the temperature reduction amplitude of the motor.

[0067] Optionally, when the iteration module 203 selects the optimal water pump speed according to the water pump speed represented by each particle and the temperature reduction amplitude of the motor, it is specifically configured to: use the following as the cost function J:

[0068] ;

[0069] Wherein, and are weights, is the water pump speed, is the motor body temperature, is the outlet water temperature; if the cost function of a particle after iteration reaches the minimum or the number of iterations reaches the threshold, the pump speed represented by the particle is taken as the optimal pump speed.

[0070] This embodiment also provides a thermal management system, including: an electronic device, a pump, a radiator, a motor, a motor controller, and a DC converter. Among them, the electronic device executes the vehicle thermal management system control method provided in the above embodiment, and is electrically connected to the pump to control the pump to reach the optimal speed. Then, predictive control for the next cycle is carried out.

[0071] Among them, referring to Figure 5 , the electronic device includes at least one processor; and a memory communicatively connected to at least one of the processors.

[0072] The memory stores instructions executable by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the above method. At least one processor in this electronic device can execute the above method, and thus has at least the same advantages as the above method.

[0073] Optionally, the electronic device further includes interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Each component is interconnected using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the electronic device, including instructions for storing graphical information in the memory or on the memory to display a GUI (Graphical User Interface) on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors and multiple memories can be used together, and / or multiple buses and multiple memories can be used together. Similarly, multiple electronic devices can be connected (for example, as a server array, a set of blade servers, or a multi-processor system), and each device provides some necessary operations. Figure 5 takes a processor 301 as an example in

[0074] The memory 302, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as program instructions / modules corresponding to the vehicle thermal management system control method in the embodiments of the present invention. The processor 301 executes various functional applications and data processing of the device by running the software programs, instructions, and modules stored in the memory 302, that is, implements the above vehicle thermal management system control method.

[0075] The memory 302 may mainly include a program storage area and a data storage area. Among them, the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 302 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some instances, the memory 302 may further include a memory remotely provided with respect to the processor 301, and these remote memories may be connected to the device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0076] The electronic device may further include: an input device 303 and an output device 304. The processor 301, the memory 302, the input device 303, and the output device 304 may be connected through a bus or other means. Figure 5 Taking the connection through the bus as an example.

[0077] The input device 303 may receive input digital or character information, and the output device 304 may include a display device, an auxiliary lighting device (for example, an LED), a tactile feedback device (for example, a vibration motor), etc. The display device may include but is not limited to a liquid crystal display (LCD), a light-emitting diode (LED) display, and a plasma display. In some embodiments, the display device may be a touch screen.

[0078] It should be understood that various forms of the processes shown above may be used, steps may be reordered, added, or deleted. For example, the steps described in the present invention may be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in the present invention can be achieved, and no limitation is imposed herein.

[0079] The above specific embodiments do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions may be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A control method for a vehicle thermal management system, characterized in that Including: Pre - train the neural network model according to the bench test data of the electric drive thermal management system; Input the parameters of the electric drive thermal management system into the neural network model to predict the outlet water temperature of the electric drive thermal management system; the parameters of the electric drive thermal management system include the water pump speed; Take the neural network model as the prediction model of model predictive control, iterate the water pump speed to obtain the optimal water pump speed; Control the water pump with the optimal water pump speed; Among them, taking the neural network model as the prediction model of model predictive control and iterating the water pump speed to obtain the optimal water pump speed includes: Use the particle swarm optimization algorithm to perform multiple iterations on the water pump speed represented by each particle, and obtain the optimized outlet water temperature through the neural network model after each iteration; Obtain the temperature reduction amplitude of the motor according to the outlet water temperature and the motor body temperature; Select the optimal water pump speed according to the water pump speed represented by each particle and the temperature reduction amplitude of the motor; Selecting the optimal water pump speed according to the water pump speed represented by each particle and the temperature reduction amplitude of the motor includes: Use the following as the cost function J: ； wherein, and are weights, is the pump speed, is the temperature of the motor body, is the outlet water temperature; If the cost function of a particle reaches the minimum after iteration or the number of iterations reaches the threshold, then use the water pump speed represented by the particle as the optimal water pump speed.

2. The method according to claim 1, characterized in that, Pre - train the neural network model according to the bench test data of the electric drive thermal management system, including: Based on the actual road condition test, install the electric drive thermal management system on the bench for testing and collect the bench test data; among them, the bench test data includes the heat generation of the electric drive system, the water pump speed, the inlet water temperature of the electric drive thermal management system, the motor body temperature, and the outlet water temperature of the electric drive thermal management system; The bench dynamometer provides road load simulation for the motor, and the environmental control system of the bench provides air intake volume and temperature control for the radiator; Use the heat generation of the electric drive system, the inlet water temperature, the motor body temperature, and the water pump speed as inputs, and the outlet water temperature as the output label to pre - train the neural network model.

3. The method according to claim 1, wherein The neural network model is a long short - term memory network (LSTM) model.

4. A control device for a vehicle thermal management system, characterized in that, Including: A training module for pre - training the neural network model according to the bench test data of the electric drive thermal management system; A prediction module for inputting the parameters of the electric drive thermal management system into the neural network model to predict the outlet water temperature of the electric drive thermal management system; the parameters of the electric drive thermal management system include the water pump speed; An iteration module for taking the neural network model as the prediction model of model predictive control and iterating the water pump speed to obtain the optimal water pump speed; A control module for controlling the water pump with the optimal water pump speed; Among them, taking the neural network model as the prediction model of model predictive control and iterating the water pump speed to obtain the optimal water pump speed includes: Use the particle swarm optimization algorithm to perform multiple iterations on the water pump speed represented by each particle, and obtain the optimized outlet water temperature through the neural network model after each iteration; Obtain the temperature reduction amplitude of the motor according to the outlet water temperature and the motor body temperature; Select the optimal water pump speed according to the water pump speed represented by each particle and the temperature reduction amplitude of the motor; Select the optimal water pump speed according to the water pump speed represented by each particle and the temperature reduction amplitude of the motor, including: Use the following as the cost function J: ； wherein, and are weights, is the pump speed, is the temperature of the motor body, is the outlet water temperature; If the cost function of a particle reaches the minimum after iteration or the number of iterations reaches the threshold, use the water pump speed represented by the particle as the optimal water pump speed.

5. A thermal management system, characterized in that, Including: An electronic device, a water pump, a radiator, a motor, a motor controller, and a DC converter; Wherein, the electronic device includes: at least one processor, and a memory communicatively connected to at least one of the processors; The memory stores instructions executable by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the vehicle thermal management system control method according to any one of claims 1-3.

Images