Control method and apparatus for battery cooling system of vehicle, and electronic device

Abstract

Disclosed in the present disclosure are a control method and apparatus for a battery cooling system of a vehicle, and an electronic device. The method relates to the field of vehicle batteries and comprises: on the basis of the current state of a vehicle, obtaining a state parameter corresponding to the current state; inputting the state parameter into a target neural network model to obtain a target control parameter of a battery cooling system; and on the basis of the target control parameter, controlling the battery cooling system of the vehicle. The present disclosure solves the technical problem of low vehicle performance caused by a poor vehicle battery temperature control effect.

Description

Method, device and electronic equipment for controlling battery cooling system of vehicle TECHNICAL FIELD

[0001] The present disclosure relates to the field of vehicles, and in particular, to a method, device and electronic equipment for controlling a battery cooling system of a vehicle. BACKGROUND

[0002] With the development of technology and the improvement of living standards, the vehicle industry is gradually developing towards electrification and intelligentization. Among them, the vehicle internal battery needs to be controlled by temperature to ensure the safety of the vehicle internal battery and the performance of the vehicle. The vehicle internal battery cooling condition needs to change in real time with the change of the power demand, and the power demand changes very frequently during the driving process of the vehicle, so the vehicle internal battery cooling demand will change at any time.

[0003] In the related art, the temperature of the vehicle internal battery is controlled by feedback signal mode, but the limitation of feedback speed will cause a time delay between receiving the signal and performing the control operation, so the temperature control of the vehicle internal battery is not always the best operation, which causes the vehicle internal battery to work in a suboptimal condition, the temperature control effect of the vehicle internal battery is poor, and thus the vehicle performance is low and the vehicle safety is poor.

[0004] At present, there is no effective solution to the above problems. SUMMARY

[0005] The embodiments of the present disclosure provide a method, device and electronic equipment for controlling a battery cooling system of a vehicle to at least solve the technical problem of low vehicle performance caused by poor vehicle battery temperature control effect.

[0006] According to an aspect of an embodiment of the present disclosure, a method for controlling a battery cooling system of a vehicle is provided, comprising: obtaining a state parameter corresponding to a current state of the vehicle based on the current state of the vehicle, wherein the state parameter comprises at least one of the following: an ambient temperature of an environment in which the vehicle is located, a driving parameter of the vehicle, a battery pack temperature, an electronic expansion valve opening degree, and a compressor speed, the battery pack, the electronic expansion valve and the compressor being located in the battery cooling system of the vehicle; inputting the state parameter into a target neural network model to obtain a target control parameter of the battery cooling system, wherein the target control parameter comprises one of the following: a target speed of the battery cooling system and a target temperature of the battery cooling system, the target speed being used to represent a predicted compressor speed of the compressor when the vehicle is in a non-driving state in a future time period, and the target temperature being used to represent a predicted refrigerant temperature at an outlet of a battery cold plate of the vehicle when the vehicle is in a driving state in the future time period, the outlet of the battery cold plate being located in the battery cooling system; and controlling the battery cooling system of the vehicle based on the target control parameter.

[0007] Further, based on the current state of the vehicle, the state parameters corresponding to the current state are obtained, including: in response to the state of the vehicle being the non-driving state, the ambient temperature and the battery pack temperature are obtained to obtain the state parameters; and in response to the state of the vehicle being the driving state, the ambient temperature, the driving parameter, the compressor speed, the electronic expansion valve opening degree and the battery pack temperature are obtained to obtain the state parameters.

[0008] Further, the target neural network model includes: a first layer neural network and a second layer neural network, the network structures of the first layer neural network and the second layer neural network are different, the state parameters are input into the target neural network model to obtain the target control parameters of the battery cooling system, including: in response to the state of the vehicle being the non-driving state, the state parameters are input into the first layer neural network to obtain the target speed of the battery cooling system; and in response to the state of the vehicle being the driving state, the state parameters are input into the second layer neural network to obtain the target temperature of the battery cooling system.

[0009] Further, based on the target control parameters, the battery cooling system of the vehicle is controlled, including: in response to the vehicle being in the non-driving state, the battery cooling system is controlled to operate based on the target speed; and in response to the vehicle being in the driving state, the electronic expansion valve opening degree or the compressor speed of the battery cooling system is adjusted based on the target temperature.

[0010] Further, based on the target temperature, the electronic expansion valve opening degree or the compressor speed of the battery cooling system is adjusted, including: the pressure of the vehicle is obtained through the pressure sensor installed on the vehicle; the refrigerant saturation temperature of the battery cooling plate outlet corresponding to the pressure is determined based on a preset corresponding relationship; the difference between the target temperature and the refrigerant saturation temperature is obtained to obtain the superheat degree; and the electronic expansion valve opening degree or the compressor speed of the battery cooling system is adjusted based on the superheat degree.

[0011] Further, after adjusting the electronic expansion valve opening degree or the compressor speed of the battery cooling system based on the superheat degree, the method further includes: in response to the value of the superheat degree being within a preset range, the adjustment of the electronic expansion valve opening degree or the compressor speed is stopped.

[0012] Further, the method further comprises: obtaining a training data set, wherein the training data set comprises: a training ambient temperature of the vehicle, a training driving parameter, a cooling fan duty cycle of the battery cooling system, a training battery pack temperature, a training compressor rotating speed, a training electronic expansion valve opening degree, and a training refrigerant temperature at the battery cooling plate outlet; inputting the cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature into an initial first layer neural network in an initial neural network model to obtain a predicted rotating speed, and inputting the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed, and the training electronic expansion valve opening degree into an initial second layer neural network in the initial neural network model to obtain a predicted temperature; training the initial first layer neural network and the initial second layer neural network based on the predicted rotating speed, the predicted temperature, the training compressor rotating speed, and the training refrigerant temperature to obtain a first layer neural network and a second layer neural network.

[0013] Further, the training of the initial first layer neural network and the initial second layer neural network based on the predicted rotating speed, the predicted temperature, the training compressor rotating speed, and the training refrigerant temperature to obtain the first layer neural network and the second layer neural network comprises: determining whether the predicted rotating speed meets the training compressor rotating speed and whether the predicted temperature meets the training refrigerant temperature; in a case where the predicted rotating speed does not meet the training compressor rotating speed or the predicted temperature does not meet the training refrigerant temperature, adjusting weights and biases of different neurons in the initial first layer neural network and the initial second layer neural network, and repeatedly inputting the cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature into the initial first layer neural network to obtain the predicted rotating speed, and inputting the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed, and the training electronic expansion valve opening degree into the initial second layer neural network to obtain the predicted temperature until the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature; and in a case where the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature, determining that the initial first layer neural network is the first layer neural network and the initial second layer neural network is the second layer neural network.

[0014] Further, the determining whether the predicted rotating speed meets the training compressor rotating speed and whether the predicted temperature meets the training refrigerant temperature comprises: obtaining a first mean square error and a first average absolute percentage error of the predicted rotating speed and the training compressor rotating speed, and obtaining a second mean square error and a second average absolute percentage error of the predicted temperature and the training refrigerant temperature; in response to the first mean square error being less than a first preset mean square error, and the first average absolute percentage error being less than a first preset percentage error, and the second mean square error being less than a second preset mean square error, and the second average absolute percentage error being less than a second preset percentage error, determining that the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature; in response to the first mean square error being greater than or equal to the first preset mean square error, or the first average absolute percentage error being greater than or equal to the first preset percentage error, or the second mean square error being greater than or equal to the second preset mean square error, or the second average absolute percentage error being greater than or equal to the second preset percentage error, determining that the predicted rotating speed does not meet the training compressor rotating speed or the predicted temperature does not meet the training refrigerant temperature.

[0015] Further, before the inputting the cooling fan duty ratio, the training ambient temperature and the training battery pack temperature into an initial first layer neural network in the initial neural network model to obtain the predicted rotating speed, and the inputting the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed and the training electronic expansion valve opening into an initial second layer neural network in the initial neural network model to obtain the predicted temperature, the method further comprises: identifying abnormal data in the training data set; determining the average value of two data adjacent to the abnormal data; replacing the abnormal data with the average value to obtain the processed training data set.

[0016] Further, the identifying abnormal data in the training data set comprises: obtaining the average value corresponding to the training data in the training data set; determining the residual error and the standard error of the training data based on the average value; determining at least one training data with the residual error greater than a preset residual error value or the standard error greater than a preset standard error value as the abnormal data.

[0017] Further, before the inputting the cooling fan duty ratio, the training ambient temperature and the training battery pack temperature into an initial first layer neural network in the initial neural network model to obtain the predicted rotating speed, and the inputting the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed and the training electronic expansion valve opening into an initial second layer neural network in the initial neural network model to obtain the predicted temperature, the method further comprises: performing normalization processing on the training data in the training data set to obtain the normalized training data set.

[0018] According to a further aspect of the embodiments of the present disclosure, a control device of a battery cooling system of a vehicle is also provided, comprising: an acquisition module configured to acquire a state parameter corresponding to a current state of the vehicle based on the current state of the vehicle, wherein the state parameter comprises at least one of the following: an ambient temperature of an environment in which the vehicle is located, a driving parameter of the vehicle, a battery pack temperature, an electronic expansion valve opening degree, and a compressor rotating speed, the battery pack, the electronic expansion valve, and the compressor being located in the battery cooling system of the vehicle; a processing module configured to input the state parameter into a target neural network model to obtain a target control parameter of the battery cooling system, wherein the target control parameter comprises one of the following: a target rotating speed of the battery cooling system and a target temperature of the battery cooling system, the target rotating speed being used to represent a predicted compressor rotating speed of the compressor when the vehicle is in a non-driving state in a future time period, and the target temperature being used to represent a predicted refrigerant temperature at a battery cold plate outlet of the vehicle when the vehicle is in a driving state in the future time period, the battery cold plate outlet being located in the battery cooling system; and a control module configured to control the battery cooling system of the vehicle based on the target control parameter.

[0019] According to a further aspect of the embodiments of the present disclosure, an electronic device is also provided, comprising: a memory storing an executable program; and a processor configured to run the program, wherein the program performs the method in the various embodiments of the present disclosure when running.

[0020] According to a further aspect of the embodiments of the present disclosure, a computer readable storage medium is also provided, comprising a stored executable program, wherein the executable program controls the device where the computer readable storage medium is located to perform the method in the various embodiments of the present disclosure when running.

[0021] According to a further aspect of the embodiments of the present disclosure, a computer program product is also provided, comprising a computer program which, when executed by a processor, implements the method in the various embodiments of the present disclosure.

[0022] In the embodiments of the present disclosure, based on a current state of a vehicle, a state parameter corresponding to the current state is acquired; the state parameter is input into a target neural network model to obtain a target control parameter of a battery cooling system; and the battery cooling system of the vehicle is controlled based on the target control parameter. Through the pre-established target neural network model, the state parameter corresponding to the current state of the vehicle is analyzed and processed to obtain the target control parameter for controlling the battery cooling system of the vehicle. The target control parameter for controlling the battery cooling system of the vehicle can be quickly and accurately determined through neural network calculation, the purpose of timely controlling the battery temperature of the vehicle as the battery temperature changes is achieved, the technical effect of improving the battery temperature control effect of the vehicle is achieved, and the technical problem of low vehicle performance caused by poor battery temperature control effect of the vehicle is solved. BRIEF DESCRIPTION OF DRAWINGS

[0023] The accompanying drawings, which are included to provide a further understanding of the disclosure and constitute a part of the disclosure, illustrate certain illustrative embodiments of the disclosure and together with the general description of the disclosure given above and the detailed description of the disclosure given below, serve to explain the disclosure. In the drawings:

[0024] FIG. 1 is a flowchart of a control method of a battery cooling system of a vehicle according to an embodiment of the disclosure;

[0025] FIG. 2 is a structural schematic diagram of a battery cooling system of a vehicle according to an embodiment of the disclosure;

[0026] FIG. 3 is a flowchart of a control method of a battery cooling system of a vehicle according to an embodiment of the disclosure;

[0027] FIG. 4 is a flowchart of a method of establishing a target neural network model according to an embodiment of the disclosure;

[0028] FIG. 5 is a structural schematic diagram of a control device of a battery cooling system of a vehicle according to an embodiment of the disclosure.

[0029] In the above drawings, the following reference signs are used: 21, condenser; 22, compressor; 23, pressure sensor; 24, second temperature sensor; 25, battery cooling plate; 26, electronic expansion valve; 27, first temperature sensor. DETAILED DESCRIPTION

[0030] In order to enable persons skilled in the art to better understand the present disclosure scheme, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by persons skilled in the art without creative labor should fall within the scope of protection of the present disclosure.

[0031] It should be noted that the terms "first", "second", and the like in the specification and claims of the present disclosure and the above drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present disclosure described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0032] Embodiment 1

[0033] According to the embodiments of the present disclosure, an embodiment of a control method of a battery cooling system of a vehicle is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in a different order.

[0034] FIG. 1 is a flowchart of an optional control method of a battery cooling system of a vehicle according to an embodiment of the present disclosure, as shown in FIG. 1, the method comprises the following steps:

[0035] In step S102, based on the current state of the vehicle, the state parameters corresponding to the current state are obtained.

[0036] Among them, the state parameters include at least one of the following: the ambient temperature of the environment in which the vehicle is located, the driving parameters of the vehicle, the battery pack temperature, the electronic expansion valve opening degree, and the compressor speed, the battery pack, the electronic expansion valve and the compressor are located in the battery cooling system of the vehicle.

[0037] The above-mentioned vehicle refers to an electric vehicle or a hybrid electric vehicle, including electric cars, electric trucks, electric buses, hybrid electric cars, hybrid electric trucks, hybrid electric buses, and other vehicles equipped with drive batteries.

[0038] The above-mentioned state parameters refer to parameters describing the state of the vehicle, and a series of data and indicators representing the state of the vehicle. Among them, the state parameters include but are not limited to the ambient temperature of the environment in which the vehicle is located, the driving parameters of the vehicle, the battery pack temperature, the electronic expansion valve opening degree, and the compressor speed.

[0039] The above-mentioned ambient temperature of the environment in which the vehicle is located refers to the ambient temperature outside the vehicle. The ambient temperature of the environment in which the vehicle is located is usually affected by factors such as changes in the weather outside the vehicle.

[0040] The above-mentioned driving parameters of the vehicle refer to the motion parameters of the vehicle, including but not limited to whether the vehicle is in a driving state, the driving speed of the vehicle, the acceleration of the vehicle, and other driving parameters describing the driving state of the vehicle.

[0041] The above-mentioned battery pack is a key component in the vehicle, which is the core of the vehicle's energy storage system, responsible for storing and providing the required electric energy to drive the electric motor, thereby realizing the driving of the vehicle. The battery pack mentioned in the present disclosure is arranged in the battery cooling system of the vehicle.

[0042] The electronic expansion valve is a device that automatically adjusts the flow rate of refrigerant. It is widely used in refrigeration, air conditioning and heat pump systems to achieve efficient operation and precise temperature control of the system. The electronic expansion valve mentioned in this disclosure refers to an electronic expansion valve configured to cool the battery inside the vehicle, which controls the temperature of the battery inside the vehicle by adjusting the flow rate of refrigerant.

[0043] The compressor is a device configured to compress refrigerant, and in this disclosure, the compressor is configured to compress refrigerant to cool the battery inside the vehicle.

[0044] In an alternative embodiment, the current state of the vehicle can be obtained by a sensor, such as driving, stopping, charging, etc. Based on the current state of the vehicle, the state parameters corresponding to the current state of the vehicle are obtained, wherein the state parameters at least include one of the following: the ambient temperature of the environment where the vehicle is located, the driving parameter of the vehicle, the battery pack temperature, the opening degree of the electronic expansion valve, and the compressor speed. Among them, the ambient temperature of the environment where the vehicle is located can be obtained by a vehicle external sensor, the driving parameter of the vehicle can be obtained by a vehicle controller, the battery pack temperature can be obtained by a vehicle battery management system, and the opening degree of the electronic expansion valve and the compressor speed can be obtained by a battery cooling system. Among them, the vehicle driving parameter includes vehicle driving speed and vehicle driving acceleration, and the battery pack temperature includes battery temperature and battery cold plate outlet refrigerant temperature. Fig. 2 is a structural schematic diagram of an alternative battery cooling system of a vehicle according to an embodiment of the disclosure, as shown in Fig. 2, the battery pack, the electronic expansion valve 26 and the compressor 22 are located in the battery cooling system of the vehicle, the first sensor 27 measures the ambient temperature of the environment where the vehicle is located, the refrigerant is compressed by the compressor 22 after passing through the condenser 21, and the battery is cooled by the battery cold plate 25. The battery pack temperature is measured by the second temperature sensor 24.

[0045] In another alternative embodiment, the current state of the vehicle can also be obtained by an on-board computer system, such as driving, stopping, charging, etc. Based on the current state of the vehicle, the state parameters corresponding to the current state of the vehicle are obtained, wherein the state parameters at least include one of the following: the ambient temperature of the environment where the vehicle is located, the driving parameter of the vehicle, the battery pack temperature, the opening degree of the electronic expansion valve, and the compressor speed. Among them, the ambient temperature of the environment where the vehicle is located can be obtained by a vehicle external sensor, and the driving parameter of the vehicle can also be obtained by an on-board diagnostic system, the battery pack temperature can be obtained by a vehicle battery management system, and the opening degree of the electronic expansion valve and the compressor speed can be obtained by a battery cooling system. Among them, the battery pack, the electronic expansion valve and the compressor are located in the battery cooling system of the vehicle.

[0046] The present disclosure can obtain more comprehensive and more relevant state parameters to the current state of the vehicle by obtaining the current state of the vehicle and obtaining corresponding state parameters through the current state of the vehicle, so that more accurate target control parameters can be obtained after analyzing the state parameters.

[0047] In step S104, the state parameters are input into the target neural network model to obtain target control parameters of the battery cooling system.

[0048] The target control parameters include one of the target speed of the battery cooling system and the target temperature of the battery cooling system, the target speed is used to represent the predicted compressor speed of the compressor in the non-driving state of the vehicle in the future time period, and the target temperature is used to represent the predicted refrigerant temperature of the battery cooling plate outlet in the driving state of the vehicle in the future time period, and the battery cooling plate outlet is located in the battery cooling system.

[0049] The target neural network model (TNNM) is a neural network architecture for specific tasks or applications. The neural network model is a brain-inspired computing model that learns and processes data by simulating the connection and information transmission between neurons. The target neural network model mentioned in the present disclosure is used to learn and calculate the state parameters of the vehicle to obtain the target control parameters of the battery of the vehicle, which is used to accurately and quickly control the battery cooling system of the vehicle. The target neural network model can be a back propagation neural network (BP neural network), a convolutional neural network (CNN), a recurrent neural network (RNN), etc.

[0050] The target control parameter refers to a parameter for controlling the battery cooling system to cool the battery so that the battery temperature is controlled within the required range, including but not limited to the target speed of the battery cooling system and the target temperature of the battery cooling system.

[0051] The target speed of the battery cooling system refers to the target speed of the compressor inside the battery cooling system, which is the required speed of the compressor when the battery temperature is controlled within the required range. In the present disclosure, the target speed of the battery cooling system is used to represent the predicted compressor speed of the compressor in the non-driving state of the vehicle in the future time period.

[0052] The target temperature of the battery cooling system refers to the temperature of the refrigerant at the outlet of the battery cooling plate in the battery cooling system, and is the required temperature of the refrigerant when the battery temperature is controlled in a required range. In the present disclosure, the target temperature of the battery cooling system is used to represent the predicted refrigerant temperature at the outlet of the battery cooling plate when the vehicle is in a driving state in a future time period.

[0053] The future time period refers to a future time period in which the target control parameter obtained through the target neural network model needs to control the battery cooling system. The future time period can be set in advance by a person or the target neural network model based on actual conditions.

[0054] The outlet of the battery cooling plate refers to the outlet end of the cooling plate arranged to cool the battery in the battery management system (BMS). The cooling plate is a device arranged to transfer the heat generated by the battery to the cooling liquid to cool the battery.

[0055] In an optional embodiment, the state parameters corresponding to the current state of the vehicle are input into the target neural network model, and the state parameters corresponding to the current state of the vehicle are analyzed and processed by the target neural network model to obtain the target control parameter of the battery cooling system. When the vehicle is in a non-driving state, the state parameters corresponding to the non-driving state of the vehicle are input into the first layer neural network in the target neural network model, and the target speed of the battery cooling system is obtained, i.e. the compressor speed of the compressor in the future time period when the vehicle is in a non-driving state is predicted by the first layer neural network. The target speed of the battery cooling system is the target control parameter of the battery cooling system when the vehicle is in a non-driving state. When the vehicle is in a driving state, the state parameters corresponding to the driving state of the vehicle are input into the second layer neural network in the target neural network model, and the target temperature of the battery cooling system is obtained, i.e. the refrigerant temperature at the outlet of the battery cooling plate when the vehicle is in a driving state in a future time period is predicted by the second layer neural network. The target temperature of the battery cooling system is the target control parameter of the battery cooling system when the vehicle is in a driving state. As shown in FIG. 2, the outlet of the battery cooling plate 25 is located in the battery cooling system.

[0056] The present disclosure analyzes and processes the state parameters corresponding to the current state of the vehicle through the pre-established target neural network model to accurately and quickly obtain the target control parameter of the battery cooling system of the vehicle, so that the battery cooling system of the vehicle can accurately and quickly control the battery temperature.

[0057] Step S106, based on the target control parameter, control the battery cooling system of the vehicle.

[0058] In an optional embodiment, after obtaining the target control parameter of the battery cooling system through the target neural network model, the battery cooling system of the vehicle is controlled based on the type of the target control parameter. When the target control parameter is the target speed, i.e., the compressor speed of the compressor in the future time period when the vehicle is in the non-driving state predicted by the first layer neural network, the compressor speed of the battery cooling system is controlled to reach the target speed. When the target control parameter is the target temperature, i.e., the refrigerant temperature at the outlet of the battery cold plate of the vehicle in the future time period when the vehicle is in the driving state predicted by the second layer neural network, the refrigerant temperature at the outlet of the battery cold plate 25 is adjusted to reach the target temperature by adjusting the opening degree of the electronic expansion valve 26 or the compressor speed of the compressor 22 in the battery cooling system.

[0059] Through the above steps, based on the current state of the vehicle, the state parameter corresponding to the current state is obtained; the state parameter is input into the target neural network model to obtain the target control parameter of the battery cooling system; and the battery cooling system of the vehicle is controlled based on the target control parameter. The present disclosure analyzes and processes the state parameter corresponding to the current state of the vehicle through the pre-established target neural network model. The present disclosure quickly outputs the target speed for controlling the battery cooling system of the vehicle when the vehicle is in the non-driving state, or quickly outputs the target temperature for controlling the battery cooling system of the vehicle when the vehicle is in the driving state, through the target speed or the target temperature to control the battery cooling system of the vehicle. The purpose of timely controlling the battery temperature of the vehicle is achieved as the battery temperature changes, thereby achieving the technical effect of improving the battery temperature control effect of the vehicle, and further solving the technical problem of low vehicle performance caused by poor battery temperature control effect of the vehicle.

[0060] Optionally, based on the current state of the vehicle, the state parameter corresponding to the current state is obtained, including: in response to the state of the vehicle being the non-driving state, obtaining the ambient temperature and the battery pack temperature to obtain the state parameter; and in response to the state of the vehicle being the driving state, obtaining the ambient temperature, the driving parameter, the compressor speed, the opening degree of the electronic expansion valve, and the battery pack temperature to obtain the state parameter.

[0061] In an optional embodiment, the state of the vehicle can be obtained by a sensor. When the vehicle is in the non-driving state, the state parameter of the vehicle in the non-driving state is obtained, wherein the state parameter of the vehicle in the non-driving state includes the ambient temperature and the battery pack temperature. When the vehicle is in the driving state, the state parameter of the vehicle in the driving state is obtained, wherein the state parameter of the vehicle in the driving state includes the ambient temperature, the driving parameter, the compressor speed, the opening degree of the electronic expansion valve, and the battery pack temperature.

[0062] In another alternative embodiment, the state of the vehicle can also be obtained by the on-board computer system as a non-driving state or a driving state. When the vehicle is in the non-driving state, state parameters of the vehicle in the non-driving state are obtained, wherein the state parameters of the vehicle in the non-driving state include the ambient temperature and the battery pack temperature. When the vehicle is in the driving state, state parameters of the vehicle in the driving state are obtained, wherein the state parameters of the vehicle in the driving state include the ambient temperature, the driving parameter, the compressor speed, the electronic expansion valve opening degree, and the battery pack temperature.

[0063] In the present disclosure, state parameters corresponding to the current state of the vehicle are obtained based on different states of the vehicle, so that when the state parameters of the vehicle are analyzed, only the state parameters influenced by the current state of the vehicle can be analyzed, thereby saving the calculation time for obtaining the target control parameter and saving energy.

[0064] Optionally, the target neural network model includes a first layer neural network and a second layer neural network, the network structures of the first layer neural network and the second layer neural network are different, and inputting the state parameters into the target neural network model to obtain the target control parameter of the battery cooling system includes: in response to the state of the vehicle being a non-driving state, inputting the state parameters into the first layer neural network to obtain the target speed of the battery cooling system; and in response to the state of the vehicle being a driving state, inputting the state parameters into the second layer neural network to obtain the target temperature of the battery cooling system.

[0065] The first layer neural network described above is a neural network configured to obtain the working parameter of the cooling system for pre-cooling the battery, i.e., a neural network configured to obtain the target speed of the battery cooling system, and the state parameters when the state of the vehicle is the non-driving state are input into the first layer neural network to obtain the target speed of the battery cooling system.

[0066] The second layer neural network described above is a neural network configured to calculate the refrigerant temperature at the outlet of the battery cooling plate, i.e., a neural network configured to obtain the target temperature of the battery cooling system, and the state parameters when the state of the vehicle is the driving state are input into the second layer neural network to obtain the target temperature of the battery cooling system.

[0067] In an alternative embodiment, the target neural network can be divided into a first layer neural network and a second layer neural network, wherein the input of the first layer neural network is the state parameters when the state of the vehicle is the non-driving state, and the output is the target speed of the battery cooling system, and the input of the second layer neural network is the state parameters when the state of the vehicle is the driving state, and the output is the target temperature of the battery cooling system. The structures of the first layer neural network and the second layer neural network are different, for example, when the target neural network model is a feedforward neural network, the number of neurons in the input layer, the hidden layer, and the output layer of the first layer neural network and the second layer neural network are different.

[0068] In the present disclosure, by dividing the target neural network into the first neural network and the second neural network, when the vehicle is in the non-driving state, the state parameters of the vehicle in the non-driving state can be analyzed and processed by the first neural network with a relatively simple structure to obtain the target rotating speed of the battery cooling system, and when the vehicle is in the driving state, the state parameters of the vehicle in the driving state can be analyzed and processed by the second neural network with a relatively complex structure to obtain the target temperature of the battery cooling system. Therefore, the target rotating speed can be quickly obtained when the vehicle is in the non-driving state, and the battery cooling system can be controlled, the vehicle battery temperature control effect is improved, and the energy consumption is saved.

[0069] Optionally, based on the target control parameter, the battery cooling system of the vehicle is controlled, including: in response to the vehicle being in the non-driving state, controlling the battery cooling system to operate based on the target rotating speed; and in response to the vehicle being in the driving state, adjusting the opening degree of the electronic expansion valve of the battery cooling system or the rotating speed of the compressor based on the target temperature.

[0070] In an optional embodiment, after the target control parameter is obtained through the target neural network model, the battery cooling system of the vehicle is controlled based on the target control parameter being the target rotating speed and the target temperature. When the vehicle is in the non-driving state, the output of the target neural network model is the target rotating speed, so the rotating speed of the compressor in the battery cooling system is controlled to be the target rotating speed; and when the vehicle is in the driving state, the output of the target neural network model is the target temperature, so the opening degree of the electronic expansion valve of the battery cooling system or the rotating speed of the compressor is adjusted to make the refrigerant temperature at the outlet of the battery cooling plate reach the target temperature.

[0071] In the present disclosure, when the vehicle is in the non-driving state, only the rotating speed of the compressor in the battery cooling system is adjusted to control the vehicle battery temperature, and when the vehicle is in the driving state, the opening degree of the electronic expansion valve of the battery cooling system or the rotating speed of the compressor is adjusted to control the vehicle battery temperature, which effectively improves the efficiency of the vehicle battery temperature control and saves energy consumption.

[0072] Optionally, based on the target temperature, the opening degree of the electronic expansion valve of the battery cooling system or the rotating speed of the compressor is adjusted, including: obtaining the pressure of the vehicle through a pressure sensor installed on the vehicle; determining the refrigerant saturation temperature at the outlet of the battery cooling plate corresponding to the pressure based on a preset corresponding relationship; obtaining the difference between the target temperature and the refrigerant saturation temperature to obtain the superheat degree; and adjusting the opening degree of the electronic expansion valve of the battery cooling system or the rotating speed of the compressor based on the superheat degree.

[0073] The pressure sensor refers to a sensor installed on the vehicle and configured to measure the pressure of the vehicle, wherein the pressure of the vehicle is the pressure of the refrigerant in the vehicle battery cooling system.

[0074] The above-mentioned refrigerant saturation temperature refers to the temperature of the refrigerant in a saturated state, that is, the temperature when the liquid and gas of the refrigerant reach an equilibrium state in the phase change process. The saturation temperature is related to the pressure, and when the pressure decreases, the saturation temperature also decreases, and when the pressure increases, the saturation temperature also increases.

[0075] The above-mentioned superheat is the difference between the target temperature and the refrigerant saturation temperature, and when the superheat is zero, the refrigerant temperature at the outlet of the battery cooling plate reaches the target temperature.

[0076] In an optional embodiment, when the vehicle is in a driving state, the pressure of the vehicle, that is, the pressure of the refrigerant, is obtained by a pressure sensor installed on the vehicle, and the saturation temperature of the refrigerant is calculated according to the relationship between the refrigerant pressure and the refrigerant saturation temperature. The superheat is obtained by subtracting the target temperature from the saturation temperature of the refrigerant, and when the superheat is greater than a preset range value, the target temperature is greater than the saturation temperature of the refrigerant, and the target temperature is equal to the saturation temperature of the refrigerant by reducing the opening degree of the electronic expansion valve of the battery cooling system or reducing the speed of the compressor. When the superheat is less than the preset range value, the target temperature is less than the saturation temperature of the refrigerant, and the target temperature is equal to the saturation temperature of the refrigerant by increasing the opening degree of the electronic expansion valve of the battery cooling system or increasing the speed of the compressor. When the superheat is within the preset range value, the opening degree of the electronic expansion valve of the battery cooling system or the speed of the compressor does not need to be adjusted.

[0077] In the present disclosure, the battery temperature is controlled by calculating the superheat and adjusting the opening degree of the electronic expansion valve of the battery cooling system or the speed of the compressor based on the superheat. The calculation process of the superheat is simple, and the control method is effective, which effectively improves the efficiency of the vehicle battery temperature control.

[0078] Optionally, after adjusting the opening degree of the electronic expansion valve of the battery cooling system or the speed of the compressor based on the superheat, the method further comprises: stopping the adjustment of the opening degree of the electronic expansion valve or the speed of the compressor in response to the value of the superheat being within the preset range value.

[0079] The above-mentioned preset range value can be determined in advance and used to determine whether to continue adjusting the electronic expansion valve or the speed of the compressor. The preset range value is usually [3, 6] degrees Celsius, but is not limited thereto, and the specific value can be set by the user according to actual test requirements, which is not limited in the present embodiment.

[0080] In an optional embodiment, when the value of the superheat is within the range of [3, 6] degrees Celsius, the battery temperature is in an optimal state, and the adjustment of the opening degree of the electronic expansion valve or the speed of the compressor is stopped, and the opening degree of the electronic expansion valve and the speed of the compressor at this time are maintained.

[0081] Optionally, the method further comprises: obtaining a training data set, wherein the training data set comprises: a training ambient temperature of the vehicle, a training driving parameter, a cooling fan duty cycle of the battery cooling system, a training battery pack temperature, a training compressor rotating speed, a training electronic expansion valve opening degree, and a training refrigerant temperature at the battery cooling plate outlet; inputting the cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature into an initial first layer neural network in an initial neural network model to obtain a predicted rotating speed, and inputting the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed, and the training electronic expansion valve opening degree into an initial second layer neural network in the initial neural network model to obtain a predicted temperature; training the initial first layer neural network and the initial second layer neural network based on the predicted rotating speed, the predicted temperature, the training compressor rotating speed, and the training refrigerant temperature to obtain a first layer neural network and a second layer neural network.

[0082] The training data set is data for training the target neural network model, including data input into the target neural network and expected output results. The training data set is one of important data for training the target neural network model. The target neural network model needs to be trained to make the target neural network model more accurate and faster after being established. The training data set can be set by human or program calculation.

[0083] The initial first layer neural network is the first layer neural network in the target neural network model just established. The initial first layer neural network is not trained and optimized, and thus is relatively rough and less accurate than the first layer neural network.

[0084] The initial second layer neural network is the second layer neural network in the target neural network model just established. The initial second layer neural network is not trained and optimized, and thus is relatively rough and less accurate than the second layer neural network.

[0085] In an optional embodiment, the training data set is set by artificial pre-setting or program calculation, and includes a training ambient temperature of the vehicle, a training driving parameter, a cooling fan duty cycle of the battery cooling system, a training battery pack temperature, a training compressor rotating speed, a training electronic expansion valve opening degree, and a training refrigerant temperature at the battery cooling plate outlet. The training ambient temperature can be an artificial simulated ambient temperature, the training driving parameter can be an artificial specified vehicle driving parameter, the cooling fan duty cycle of the battery cooling system is determined according to the working mode of the cooling fan of the battery cooling system, and the training battery pack temperature, the training compressor rotating speed, the training electronic expansion valve opening degree, and the training refrigerant temperature at the battery cooling plate outlet are determined according to artificial demand simulation. The cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature are input into an initial first layer neural network in an initial neural network model to obtain a predicted rotating speed, and the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed, and the training electronic expansion valve opening degree are input into an initial second layer neural network in the initial neural network model to obtain a predicted temperature. The initial first layer neural network and the initial second layer neural network are trained through the predicted rotating speed, the predicted temperature, the training compressor rotating speed, and the training refrigerant temperature, and the predicted rotating speed is adjusted to meet the training compressor rotating speed and the predicted temperature is adjusted to meet the training refrigerant temperature by adjusting the neuron weight and the like. When the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature, the training of the initial first layer neural network and the initial second layer neural network is completed, and the initial first layer neural network at this time is the first layer neural network and the initial second layer neural network is the second layer neural network.

[0086] In the present disclosure, the initial first layer neural network and the initial second layer neural network are trained through the training data set to obtain more accurate first layer neural network and second layer neural network, so that the output target rotating speed and target temperature are more accurate, and the vehicle battery temperature is accurately controlled.

[0087] Optionally, based on the predicted rotating speed, the predicted temperature, the training compressor rotating speed and the training refrigerant temperature, the initial first-layer neural network and the initial second-layer neural network are trained to obtain the first-layer neural network and the second-layer neural network, including: determining whether the predicted rotating speed meets the training compressor rotating speed and whether the predicted temperature meets the training refrigerant temperature; in the case that the predicted rotating speed does not meet the training compressor rotating speed or the predicted temperature does not meet the training refrigerant temperature, adjusting the weights and biases of different neurons in the initial first-layer neural network and the initial second-layer neural network, and repeatedly inputting the cooling fan duty ratio, the training ambient temperature and the training battery pack temperature into the initial first-layer neural network to obtain the predicted rotating speed, and simultaneously inputting the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed and the training electronic expansion valve opening into the initial second-layer neural network to obtain the predicted temperature until the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature; in the case that the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature, determining that the initial first-layer neural network is the first-layer neural network and the initial second-layer neural network is the second-layer neural network.

[0088] The above-mentioned neuron is a basic unit of a neural network. Each neuron receives input signals from other neurons, processes them, and then generates an output signal that can be passed to other neurons.

[0089] The above-mentioned neuron weight refers to the weight of each input signal in the neuron, which determines the degree of influence of the input signal on the neuron output. The weight is a parameter learned by the neural network during training.

[0090] The above-mentioned neuron bias refers to the activation threshold of the neuron, which is a constant term. The above-mentioned neuron bias is a parameter learned by the neural network during training.

[0091] In an optional embodiment, after the preset rotating speed is obtained, the preset rotating speed is compared with the training compressor rotating speed, and after the preset temperature is obtained, the preset temperature is compared with the training refrigerant temperature. When the predicted rotating speed does not meet the training compressor rotating speed, or the predicted temperature does not meet the training refrigerant temperature, the weights and biases of different neurons in the initial first-layer neural network and the initial second-layer neural network are adjusted, and the steps of inputting the cooling fan duty ratio, the training ambient temperature and the training battery pack temperature into the initial first-layer neural network to obtain the predicted rotating speed, and inputting the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed and the training electronic expansion valve opening degree into the initial second-layer neural network to obtain the predicted temperature are repeated until the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature. When the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature, the training of the initial first-layer neural network and the initial second-layer neural network is completed, and at this time, the initial first-layer neural network is determined as the first-layer neural network, and the initial second-layer neural network is determined as the second-layer neural network.

[0092] In the present disclosure, by adjusting the neuron weights and biases and continuously obtaining the predicted rotating speed and the predicted temperature, the initial first-layer neural network and the initial second-layer neural network are continuously optimized, and whether the optimization is completed is judged by whether the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature, so as to obtain more accurate first-layer neural network and second-layer neural network.

[0093] Optionally, determining whether the predicted rotating speed meets the training compressor rotating speed and whether the predicted temperature meets the training refrigerant temperature comprises: obtaining a first mean square error and a first average absolute percentage error of the predicted rotating speed and the training compressor rotating speed, and obtaining a second mean square error and a second average absolute percentage error of the predicted temperature and the training refrigerant temperature; in response to the first mean square error being less than a first preset mean square error, the first average absolute percentage error being less than a first preset percentage error, the second mean square error being less than a second preset mean square error, and the second average absolute percentage error being less than a second preset percentage error, it is determined that the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature; in response to the first mean square error being greater than or equal to the first preset mean square error, or the first average absolute percentage error being greater than or equal to the first preset percentage error, or the second mean square error being greater than or equal to the second preset mean square error, or the second average absolute percentage error being greater than or equal to the second preset percentage error, it is determined that the predicted rotating speed does not meet the training compressor rotating speed, or the predicted temperature does not meet the training refrigerant temperature.

[0094] In an optional embodiment, the fitting degree of the predicted rotating speed and the training compressor rotating speed is judged by the first mean square error and the first average absolute percentage error, and the fitting degree of the predicted temperature and the training refrigerant temperature is judged by the second mean square error and the second average absolute percentage error. The first mean square error and the first average absolute percentage error of the predicted rotating speed and the training compressor rotating speed are obtained by calculation, and the second mean square error and the second average absolute percentage error of the predicted temperature and the training refrigerant temperature are obtained. When the first mean square error is less than the first preset mean square error, and the first average absolute percentage error is less than the first preset percentage error, it is determined that the predicted rotating speed meets the training compressor rotating speed; when the second mean square error is less than the second preset mean square error, and the second average absolute percentage error is less than the second preset percentage error, it is determined that the predicted temperature meets the training refrigerant temperature. When the first mean square error is greater than or equal to the first preset mean square error, or the first average absolute percentage error is greater than or equal to the first preset percentage error, it is determined that the predicted rotating speed does not meet the training compressor rotating speed; when the second mean square error is greater than or equal to the second preset mean square error, or the second average absolute percentage error is greater than or equal to the second preset percentage error, it is determined that the predicted temperature does not meet the training refrigerant temperature.

[0095] The present disclosure judges the fitting of the predicted rotating speed and the training compressor rotating speed by the first mean square error and the first average absolute percentage error of the predicted rotating speed and the training compressor rotating speed, and judges the fitting of the predicted temperature and the training refrigerant temperature by the second mean square error and the second average absolute percentage error of the predicted temperature and the training refrigerant temperature, so that the fitting of the predicted rotating speed and the training compressor rotating speed is more accurate, the fitting of the predicted temperature and the training refrigerant temperature is more accurate, and the accuracy of the vehicle battery temperature control is effectively improved.

[0096] Optionally, before the cooling fan duty ratio, the training environment temperature and the training battery pack temperature are input into the initial first layer neural network in the initial neural network model to obtain the predicted rotating speed, and the training battery pack temperature, the training environment temperature, the training driving parameter, the training compressor rotating speed and the training electronic expansion valve opening are input into the initial second layer neural network in the initial neural network model to obtain the predicted temperature, the method further comprises: identifying abnormal data in the training data set; determining the average value of two data adjacent to the abnormal data; replacing the abnormal data with the average value to obtain the processed training data set.

[0097] The above-mentioned abnormal data refers to data with large errors in the training data set, and the cause of the abnormal data may be human negligence or program calculation vulnerability, and the abnormal data needs to be eliminated to ensure the effectiveness of the training of the initial first layer neural network and the initial second layer neural network.

[0098] The replacing of the abnormal data by the average value refers to putting the average value of the two data adjacent to the abnormal data into the training data set after the abnormal data is removed.

[0099] In an optional embodiment, before the training data set is input into the initial first layer neural network and the initial second layer neural network, the abnormal data in the training data set is identified by the residual error of the training data and the standard error, and the average value of the two data adjacent to the abnormal data is calculated, the abnormal data is replaced by the average value of the two data adjacent to the abnormal data, and the processed training data set is obtained.

[0100] In the present disclosure, by removing the abnormal data in the training data set, the training of the initial first layer neural network and the initial second layer neural network is more accurate, and the calculation of the first layer neural network and the second layer neural network on the target rotating speed and the target temperature is more accurate.

[0101] Optionally, the identification of the abnormal data in the training data set comprises: obtaining the average value corresponding to the training data in the training data set; determining the residual error and the standard error of the training data based on the average value; and determining at least one training data with a residual error greater than a preset residual error value or a standard error greater than a preset standard error value as the abnormal data.

[0102] In an optional embodiment, the training data x1, x2, x3, …, xn in the training data set are selected n , and the average value corresponding to the training data is calculated , wherein the average value corresponding to the training data is calculated according to the following formula:

[0103] After the average value corresponding to the training data is obtained , the residual error of the training data is calculated, wherein the calculation formula of the residual error of the training data is as follows:

[0104] , wherein v i is the residual error of the training data, x i is any one training data, , and the average value corresponding to the training data is

[0105] The standard error of the training data is calculated by the Bessel formula, and the calculation formula of the standard error of the training data is as follows:

[0106] , wherein σ is the standard error of the training data, x i is any one training data, , and the average value corresponding to the training data is

[0107] When the residual error of any one training data is greater than the preset residual error value, or the standard error is greater than the preset standard error value, the training data is determined as abnormal data.

[0108] In the present disclosure, by comparing the residual error of the training data with the preset residual error value and comparing the standard error with the preset standard error value, the abnormal data in the training data is more accurately judged, and the normal training data is avoided to be eliminated when eliminating the abnormal data.

[0109] Optionally, the cooling fan duty ratio, the training environment temperature, and the training battery pack temperature are input into an initial first layer neural network in the initial neural network model to obtain the predicted rotating speed, and the training battery pack temperature, the training environment temperature, the training driving parameter, the training compressor rotating speed, and the training electronic expansion valve opening degree are input into an initial second layer neural network in the initial neural network model to obtain the predicted temperature, and before that, the method further includes: performing normalization processing on the training data in the training data set to obtain a normalized training data set.

[0110] In an optional embodiment, before the training data set is input into the initial first layer neural network and the initial second layer neural network, the training data in the training data set is also normalized, and the calculation formula for normalizing the training data in the training data set is as follows:

[0111] Wherein, y i is the training data in the normalized training data set, x i is any one training data, x min is the minimum value in the training data, x max is the maximum value in the training data, i = 1, 2, 3, …, n.

[0112] The normalized training data set y is a set of training data in the normalized training data set. y = {y1, y2, y3, …, y n}.

[0113] By normalizing the training data in the training data set, the present disclosure can accelerate the training speed of the initial first layer neural network and the initial second layer neural network, and improve the performance of the initial first layer neural network and the initial second layer neural network.

[0114] Optionally, the network structure of the initial first layer neural network and the initial second layer neural network is determined based on an empirical formula.

[0115] In an optional embodiment, for example, when the target neural network model is a feedforward neural network, the network structure of the initial first-layer neural network and the initial second-layer neural network is obtained through an empirical formula, where the empirical formula is as follows:

[0116] wherein, l is the number of hidden layer nodes in the feedforward neural network, n is the number of input layer nodes in the feedforward neural network, m is the number of output layer nodes in the feedforward neural network, and a is a constant between 0 and 10, wherein the value of a is artificially set according to actual conditions.

[0117] The present disclosure determines the network structure of the initial first-layer neural network and the initial second-layer neural network through an empirical formula, so that the training speed of the initial first-layer neural network and the initial second-layer neural network can be improved when the initial first-layer neural network and the initial second-layer neural network are trained.

[0118] A preferred embodiment of the present disclosure will be described in detail below in combination with FIG. 3, which is a flowchart of a control method of a battery cooling system of a vehicle according to a preferred embodiment of the present disclosure. As shown in FIG. 3, the control method of the battery cooling system of the vehicle includes the following steps:

[0119] In step S301, a state parameter corresponding to a current state of the vehicle is obtained.

[0120] In an optional embodiment, the state parameter corresponding to the current state of the vehicle is obtained based on the current state of the vehicle, wherein the state parameter at least includes one of the following: an ambient temperature of an environment in which the vehicle is located, a driving parameter of the vehicle, a battery pack temperature, an opening degree of an electronic expansion valve, and a compressor rotating speed.

[0121] In step S302, it is determined whether the vehicle is currently in a driving state.

[0122] In an optional embodiment, when the vehicle is currently in the driving state, step S305 is performed, and when the vehicle is not in the driving state but in a non-driving state, step S303 is performed. The present disclosure obtains the state parameter corresponding to the current state of the vehicle based on different states of the vehicle, so that when the state parameter of the vehicle is analyzed, only the state parameter affected by the current state of the vehicle can be analyzed, thereby saving the calculation time for obtaining the target control parameter and saving energy.

[0123] In step S303, the state parameter is input into the first-layer neural network.

[0124] In an optional embodiment, after it is determined that the vehicle is not in the driving state but in the non-driving state, the ambient temperature of the environment in which the vehicle is located, the battery pack temperature, and the compressor rotating speed are input into the first-layer neural network to obtain a target rotating speed of the battery cooling system, and step S304 is performed.

[0125] Step S304, output the target rotating speed, and control the rotating speed of the compressor by the target rotating speed.

[0126] In an optional embodiment, when the vehicle is in the non-driving state, the output of the target neural network model is the target rotating speed, and thus the rotating speed of the compressor in the battery cooling system is controlled to the target rotating speed to control the temperature of the vehicle battery.

[0127] Step S305, input the state parameters into the second layer neural network.

[0128] In an optional embodiment, after it is judged that the vehicle is in the driving state, the ambient temperature of the environment in which the vehicle is located, the driving parameters of the vehicle, the battery pack temperature, the opening degree of the electronic expansion valve, and the rotating speed of the compressor are input into the second layer neural network to obtain the target temperature of the battery cooling system, and step S306 is performed.

[0129] Step S306, output the target temperature, and calculate the superheat degree.

[0130] In an optional embodiment, after it is judged that the vehicle is in the driving state, the state parameters are input into the second layer neural network, the target temperature is output by the second layer neural network, the pressure of the vehicle, i.e., the pressure of the refrigerant, is obtained by the pressure sensor installed on the vehicle, the saturation temperature of the refrigerant is calculated according to the relationship between the refrigerant pressure and the refrigerant saturation temperature, and the superheat degree is obtained by subtracting the target temperature from the saturation temperature of the refrigerant.

[0131] Step S307, adjust the opening degree of the electronic expansion valve or the rotating speed of the compressor based on the superheat degree.

[0132] In an optional embodiment, when the superheat degree is positive, the target temperature is greater than the saturation temperature of the refrigerant, the opening degree of the electronic expansion valve of the battery cooling system is reduced, or the rotating speed of the compressor is reduced so that the target temperature is equal to the saturation temperature of the refrigerant; when the superheat degree is negative, the target temperature is less than the saturation temperature of the refrigerant, the opening degree of the electronic expansion valve of the battery cooling system is increased, or the rotating speed of the compressor is increased so that the target temperature is equal to the saturation temperature of the refrigerant. When the superheat degree is zero, the opening degree of the electronic expansion valve of the battery cooling system does not need to be adjusted, or the rotating speed of the compressor does not need to be adjusted.

[0133] The establishment of the target neural network model in the present disclosure will be described in detail below in combination with FIG. 4, which is a flowchart of an optional method for establishing a target neural network model according to an embodiment of the present disclosure. The method for establishing the target neural network model includes the following steps:

[0134] Step S401, establish an initial first layer neural network and an initial second layer neural network.

[0135] In an optional embodiment, the network structure of the initial first-layer neural network and the initial second-layer neural network is determined based on an empirical formula to obtain the initial first-layer neural network and the initial second-layer neural network.

[0136] In step S402, a training data set is obtained.

[0137] In an optional embodiment, the training data set is set by manual pre-setting or program calculation, and the training data set includes a training ambient temperature of the vehicle, a training driving parameter, a cooling fan duty cycle of the battery cooling system, a training battery pack temperature, a training compressor rotating speed, a training electronic expansion valve opening degree, and a training refrigerant temperature at the battery cooling plate outlet. The training ambient temperature can be a manually simulated ambient temperature, the training driving parameter can be a manually specified vehicle driving parameter, the cooling fan duty cycle of the battery cooling system is determined according to the working mode of the cooling fan of the battery cooling system, and the training battery pack temperature, the training compressor rotating speed, the training electronic expansion valve opening degree, and the training refrigerant temperature at the battery cooling plate outlet are determined according to manual demand simulation.

[0138] In step S403, abnormal data in the training data set is obtained and replaced to obtain a processed training data set.

[0139] In an optional embodiment, before the training data set is input into the initial first-layer neural network and the initial second-layer neural network, the abnormal data in the training data set is identified by the residual error and the standard error of the training data, and the average value of two data adjacent to the abnormal data is calculated. The average value of the two data adjacent to the abnormal data is used to replace the abnormal data to obtain the processed training data set.

[0140] In step S404, the processed training data set is normalized to obtain a normalized training data set.

[0141] In an optional embodiment, after the processed training data set is obtained, the processed training data set is normalized to accelerate the training speed of the initial first-layer neural network and the initial second-layer neural network, improve the performance of the initial first-layer neural network and the initial second-layer neural network, and obtain the normalized training data set.

[0142] In step S405, the normalized training data set is input into the initial first-layer neural network and the initial second-layer neural network to obtain a predicted rotating speed and a predicted temperature.

[0143] In an alternative embodiment, the cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature are input into an initial first layer neural network in the initial neural network model to obtain the predicted rotational speed, and the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotational speed, and the training electronic expansion valve opening degree are input into an initial second layer neural network in the initial neural network model to obtain the predicted temperature.

[0144] In step S406, it is determined whether the predicted rotational speed and the predicted temperature both meet the requirements.

[0145] In an alternative embodiment, after obtaining the predicted rotational speed and the predicted temperature, the first mean square error and the first average absolute percentage error between the predicted rotational speed and the training compressor rotational speed are obtained by calculation, and the second mean square error and the second average absolute percentage error between the predicted temperature and the training refrigerant temperature are obtained. When the first mean square error is less than a first preset mean square error, and the first average absolute percentage error is less than a first preset percentage error, it is determined that the predicted rotational speed meets the training compressor rotational speed, i.e., the predicted rotational speed meets the requirements. When the second mean square error is less than a second preset mean square error, and the second average absolute percentage error is less than a second preset percentage error, it is determined that the predicted temperature meets the training refrigerant temperature, i.e., the predicted temperature meets the requirements. When the first mean square error is greater than or equal to the first preset mean square error, or the first average absolute percentage error is greater than or equal to the first preset percentage error, it is determined that the predicted rotational speed does not meet the training compressor rotational speed, i.e., the predicted rotational speed does not meet the requirements. When the second mean square error is greater than or equal to the second preset mean square error, or the second average absolute percentage error is greater than or equal to the second preset percentage error, it is determined that the predicted temperature does not meet the training refrigerant temperature, i.e., the predicted temperature does not meet the requirements. When the predicted rotational speed and the predicted temperature both meet the requirements, step S408 is performed, otherwise step S407 is performed.

[0146] In step S407, the weights and biases of different neurons are adjusted.

[0147] In an alternative embodiment, when the predicted rotational speed does not meet the training compressor rotational speed, or the predicted temperature does not meet the training refrigerant temperature, the weights and biases of different neurons in the initial first layer neural network and the initial second layer neural network are adjusted, the initial first layer neural network and the initial second layer neural network are optimized, and then step S405 is performed. The steps of inputting the cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature into the initial first layer neural network to obtain the predicted rotational speed, and inputting the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotational speed, and the training electronic expansion valve opening degree into the initial second layer neural network to obtain the predicted temperature are repeated until the predicted rotational speed meets the training compressor rotational speed, and the predicted temperature meets the training refrigerant temperature.

[0148] In step S408, the first neural network and the second neural network are obtained.

[0149] In an optional embodiment, when the predicted rotation speed meets the training compressor rotation speed and the predicted temperature meets the training refrigerant temperature, the training of the initial first-layer neural network and the initial second-layer neural network is completed, and at this time, the initial first-layer neural network is determined as the first-layer neural network and the initial second-layer neural network is determined as the second-layer neural network.

[0150] Embodiment 2

[0151] According to another aspect of the embodiments of the present disclosure, a control device of a battery cooling system of a vehicle is also provided, which can perform the control method of the battery cooling system of the vehicle provided in Embodiment 1 above, and the specific implementation and preferred application scenarios are the same as those of Embodiment 1 above, which will not be repeated here.

[0152] FIG. 5 is a structural schematic diagram of an optional control device of a battery cooling system of a vehicle according to an embodiment of the present disclosure. As shown in FIG. 5, the device includes: an acquisition module 50 configured to acquire state parameters corresponding to a current state of the vehicle based on the current state of the vehicle, wherein the state parameters include at least one of the following: an ambient temperature of an environment in which the vehicle is located, a driving parameter of the vehicle, a battery pack temperature, an electronic expansion valve opening degree, and a compressor rotation speed, wherein the battery pack, the electronic expansion valve, and the compressor are located in the battery cooling system of the vehicle; a processing module 52 configured to input the state parameters into a target neural network model to obtain a target control parameter of the battery cooling system, wherein the target control parameter includes one of the following: a target rotation speed of the battery cooling system and a target temperature of the battery cooling system, the target rotation speed being used to represent a predicted compressor rotation speed of the compressor when the vehicle is in a non-driving state in a future time period, and the target temperature being used to represent a predicted refrigerant temperature at an outlet of a battery cold plate of the vehicle when the vehicle is in a driving state in the future time period, the outlet of the battery cold plate being located in the battery cooling system; and a control module 54 configured to control the battery cooling system of the vehicle based on the target control parameter.

[0153] Optionally, the acquisition module includes: a first acquisition unit configured to acquire the ambient temperature and the battery pack temperature to obtain the state parameters in response to the state of the vehicle being the non-driving state; and a second acquisition unit configured to acquire the ambient temperature, the driving parameter, the compressor rotation speed, the electronic expansion valve opening degree, and the battery pack temperature to obtain the state parameters in response to the state of the vehicle being the driving state.

[0154] Optionally, the target neural network model comprises a first layer neural network and a second layer neural network, the first layer neural network and the second layer neural network have different network structures, and the processing module comprises a first processing unit configured to input the state parameter into the first layer neural network to obtain the target rotating speed of the battery cooling system in response to the state of the vehicle being the non-driving state, and a second processing unit configured to input the state parameter into the second layer neural network to obtain the target temperature of the battery cooling system in response to the state of the vehicle being the driving state.

[0155] Optionally, the control module comprises a first control unit configured to control the battery cooling system to operate based on the target rotating speed in response to the vehicle being in the non-driving state, and a second control unit configured to adjust the opening degree of the electronic expansion valve or the rotating speed of the compressor of the battery cooling system based on the target temperature in response to the vehicle being in the driving state.

[0156] Optionally, the second control unit comprises an acquisition subunit configured to acquire the pressure of the vehicle through a pressure sensor installed on the vehicle, a determination subunit configured to determine the refrigerant saturation temperature of the battery cooling plate outlet corresponding to the pressure based on a preset corresponding relationship, a difference subunit configured to obtain a difference between the target temperature and the refrigerant saturation temperature to obtain an overheating degree, and an adjustment subunit configured to adjust the opening degree of the electronic expansion valve or the rotating speed of the compressor of the battery cooling system based on the overheating degree.

[0157] Optionally, the adjustment subunit is further configured to stop adjusting the opening degree of the electronic expansion valve or the rotating speed of the compressor in response to the value of the overheating degree being within a preset range.

[0158] Optionally, the device further comprises an establishment module configured to acquire a training data set, wherein the training data set comprises a training ambient temperature of the vehicle, a training driving parameter, a cooling fan duty cycle of the battery cooling system, a training battery pack temperature, a training compressor rotating speed, a training electronic expansion valve opening degree, and a training refrigerant temperature of the battery cooling plate outlet, a prediction module configured to input the cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature into an initial first layer neural network in an initial neural network model to obtain a predicted rotating speed, and input the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed, and the training electronic expansion valve opening degree into an initial second layer neural network in the initial neural network model to obtain a predicted temperature, and a determination module configured to train the initial first layer neural network and the initial second layer neural network based on the predicted rotating speed, the predicted temperature, the training compressor rotating speed, and the training refrigerant temperature to obtain the first layer neural network and the second layer neural network.

[0159] Optionally, the determining module comprises: a judging unit configured to determine whether the predicted rotating speed meets the training compressor rotating speed and whether the predicted temperature meets the training refrigerant temperature; a circulating unit configured to, in the case that the predicted rotating speed does not meet the training compressor rotating speed or the predicted temperature does not meet the training refrigerant temperature, adjust the weights and biases of different neurons in the initial first-layer neural network and the initial second-layer neural network, and repeatedly input the cooling fan duty ratio, the training ambient temperature and the training battery pack temperature into the initial first-layer neural network to obtain the predicted rotating speed, and simultaneously input the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed and the training electronic expansion valve opening into the initial second-layer neural network to obtain the predicted temperature, until the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature; and a determining unit configured to, in the case that the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature, determine that the initial first-layer neural network is the first-layer neural network and the initial second-layer neural network is the second-layer neural network.

[0160] Optionally, the judging unit comprises: a calculating sub-unit configured to obtain a first mean square error and a first average absolute percentage error of the predicted rotating speed and the training compressor rotating speed, and obtain a second mean square error and a second average absolute percentage error of the predicted temperature and the training refrigerant temperature; a comparing sub-unit configured to, in response to the first mean square error being less than a first preset mean square error, the first average absolute percentage error being less than a first preset percentage error, the second mean square error being less than a second preset mean square error, and the second average absolute percentage error being less than a second preset percentage error, determine that the predicted rotating speed meets the training compressor rotating speed and the predicted temperature meets the training refrigerant temperature; and a predicting sub-unit configured to, in response to the first mean square error being greater than or equal to the first preset mean square error, the first average absolute percentage error being greater than or equal to the first preset percentage error, the second mean square error being greater than or equal to the second preset mean square error, or the second average absolute percentage error being greater than or equal to the second preset percentage error, determine that the predicted rotating speed does not meet the training compressor rotating speed or the predicted temperature does not meet the training refrigerant temperature.

[0161] Optionally, before the cooling fan duty ratio, the training ambient temperature and the training battery pack temperature are input into the initial first-layer neural network in the initial neural network model to obtain the predicted rotating speed, and simultaneously the training battery pack temperature, the training ambient temperature, the training driving parameter, the training compressor rotating speed and the training electronic expansion valve opening are input into the initial second-layer neural network in the initial neural network model to obtain the predicted temperature, the device further comprises: an anomaly identifying module configured to identify abnormal data in the training data set; an average value module configured to determine an average value of two data adjacent to the abnormal data; and a replacing module configured to replace the abnormal data with the average value to obtain a processed training data set.

[0162] Optionally, the anomaly identification module comprises: an average unit configured to obtain the average value of the training data in the training data set; the anomaly identification module is configured to determine the residual error and the standard error of the training data based on the average value; and the anomaly determination unit is configured to determine that at least one training data with the residual error greater than a preset residual error value or the standard error greater than a preset standard error value is the abnormal data.

[0163] Optionally, the cooling fan duty ratio, the training environment temperature, and the training battery pack temperature are input into an initial first layer neural network in the initial neural network model to obtain the predicted rotating speed, and the training battery pack temperature, the training environment temperature, the training driving parameter, the training compressor rotating speed, and the training electronic expansion valve opening degree are input into an initial second layer neural network in the initial neural network model to obtain the predicted temperature, and before that, the device further comprises a normalization module configured to perform normalization processing on the training data in the training data set to obtain a normalized training data set.

[0164] Embodiment 3

[0165] Embodiments of the present disclosure also provide an electronic device, comprising: a memory storing an executable program; and a processor configured to execute the program, wherein the program performs the method in the embodiments of the present disclosure when executed.

[0166] Embodiment 4

[0167] Embodiments of the present disclosure also provide a computer-readable storage medium comprising a stored executable program, wherein the computer-readable storage medium controls the device where the computer-readable storage medium is located to perform the method in the embodiments of the present disclosure when the executable program is executed.

[0168] Embodiment 5

[0169] Embodiments of the present disclosure also provide a computer program product comprising a computer program, wherein the computer program implements the method in the embodiments of the present disclosure when executed by a processor.

[0170] Embodiment 6

[0171] Embodiments of the present disclosure also provide a computer program, wherein the computer program implements the method in the embodiments of the present disclosure when executed by a processor.

[0172] The above-mentioned serial numbers of the embodiments of the present disclosure are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0173] In the above-mentioned embodiments of the present disclosure, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0174] In several embodiments provided by the present disclosure, it should be understood that the disclosed technology can be implemented in other manners. For example, the described unit embodiments can be divided into other ways, and the functions of the units can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be implemented by using some interfaces, and the indirect couplings or communication connections can be implemented in electronic, mechanical, or other forms.

[0175] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, can be located in one place, or can be distributed on multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0176] In addition, each functional unit in the various embodiments of the present disclosure can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be implemented in the form of hardware, or in the form of a software functional unit.

[0177] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the present disclosure essentially or substantially, or all or part of the technical solutions, can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present disclosure. The foregoing storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, and various other media that can store program codes.

[0178] The above descriptions are only preferred embodiments of the present disclosure, and it should be pointed out that, for those skilled in the art, without departing from the principles of the present disclosure, several improvements and refinements can be made, and these improvements and refinements should be considered as the protection scope of the present disclosure. Industrial applicability

[0179] In the embodiments of the present disclosure, based on the current state of the vehicle, a state parameter corresponding to the current state is acquired; the state parameter is input into a target neural network model to obtain a target control parameter of the battery cooling system; and the battery cooling system of the vehicle is controlled based on the target control parameter. Through the target neural network model established in advance, the state parameter corresponding to the current state of the vehicle is analyzed and processed to obtain the target control parameter for controlling the battery cooling system of the vehicle. The target control parameter for controlling the battery cooling system of the vehicle can be quickly and accurately determined through neural network calculation, the purpose of timely controlling the battery temperature of the vehicle as the battery temperature changes is achieved, the technical effect of improving the battery temperature control effect of the vehicle is achieved, and the technical problem of low vehicle performance caused by poor battery temperature control effect of the vehicle is solved.

Claims

1. A control method for a vehicle battery cooling system, comprising: Based on the current state of the vehicle, obtain the state parameters corresponding to the current state, wherein the state parameters include at least one of the following: the ambient temperature of the environment in which the vehicle is located, the driving parameters of the vehicle, the battery pack temperature, the opening degree of the electronic expansion valve, and the compressor speed, wherein the battery pack, the electronic expansion valve and the compressor are located in the battery cooling system of the vehicle; The state parameters are input into the target neural network model to obtain the target control parameters of the battery cooling system. The target control parameters include one of the following: the target rotational speed of the battery cooling system and the target temperature of the battery cooling system. The target rotational speed is used to characterize the predicted compressor rotational speed when the vehicle is in a non-driving state in the future time period. The target temperature is used to characterize the predicted refrigerant temperature at the battery cold plate outlet when the vehicle is in a driving state in the future time period. The battery cold plate outlet is located in the battery cooling system. The vehicle's battery cooling system is controlled based on the target control parameters.

2. The method according to claim 1, wherein, Based on the vehicle's current state, obtain the state parameters corresponding to the current state, including: In response to the vehicle being in a non-driving state, the ambient temperature and the battery pack temperature are acquired to obtain the state parameters; In response to the vehicle being in a driving state, the ambient temperature, driving parameters, compressor speed, electronic expansion valve opening, and battery pack temperature are acquired to obtain the state parameters.

3. The method according to claim 1, wherein, The target neural network model includes: a first-layer neural network and a second-layer neural network. The network structures of the first-layer neural network and the second-layer neural network are different. The state parameters are input into the target neural network model to obtain the target control parameters of the battery cooling system, including: In response to the vehicle being in a non-driving state, the state parameters are input into the first layer of the neural network to obtain the target rotational speed of the battery cooling system; In response to the vehicle being in a driving state, the state parameters are input into the second layer of the neural network to obtain the target temperature of the battery cooling system.

4. The method according to claim 1, wherein, Based on the target control parameters, the vehicle's battery cooling system is controlled, including: In response to the vehicle being in a stationary state, the battery cooling system is controlled to operate based on the target rotational speed; In response to the vehicle being in motion, the opening of the electronic expansion valve of the battery cooling system or the speed of the compressor is adjusted based on the target temperature.

5. The method according to claim 4, wherein, Based on the target temperature, adjust the opening of the electronic expansion valve of the battery cooling system, or the compressor speed, including: The pressure of the vehicle is obtained by a pressure sensor installed on the vehicle. Based on a preset correspondence, the refrigerant saturation temperature at the outlet of the battery cold plate corresponding to the pressure is determined; The difference between the target temperature and the refrigerant saturation temperature is obtained to determine the superheat. The opening degree of the electronic expansion valve of the battery cooling system or the compressor speed is adjusted based on the superheat.

6. The method according to claim 5, wherein, The method further includes adjusting the opening of the electronic expansion valve of the battery cooling system based on the superheat, or, after adjusting the compressor speed: In response to the superheat value being within a preset range, the adjustment of the opening of the electronic expansion valve or the speed of the compressor is stopped.

7. The method according to claim 1, wherein, The method further includes: Obtain a training dataset, wherein the training dataset includes: the training ambient temperature of the vehicle, training driving parameters, the duty cycle of the cooling fan of the battery cooling system, the training battery pack temperature, the training compressor speed, the training electronic expansion valve opening, and the training refrigerant temperature at the outlet of the battery cold plate. The cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature are input into the initial first layer of the initial neural network model to obtain the predicted rotational speed. At the same time, the training battery pack temperature, the training ambient temperature, the training driving parameters, the training compressor speed, and the training electronic expansion valve opening are input into the initial second layer of the initial neural network model to obtain the predicted temperature. Based on the predicted rotational speed, the predicted temperature, the training compressor rotational speed, and the training refrigerant temperature, the initial first-layer neural network and the initial second-layer neural network are trained to obtain the first-layer neural network and the second-layer neural network.

8. The method according to claim 7, wherein, Based on the predicted rotational speed, the predicted temperature, the training compressor rotational speed, and the training refrigerant temperature, the initial first-layer neural network and the initial second-layer neural network are trained to obtain the first-layer neural network and the second-layer neural network, including: Determine whether the predicted rotational speed meets the training compressor rotational speed, and whether the predicted temperature meets the training refrigerant temperature; If the predicted rotational speed does not meet the training compressor speed, or the predicted temperature does not meet the training refrigerant temperature, the weights and biases of different neurons in the initial first layer neural network and the initial second layer neural network are adjusted. The cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature are repeatedly input into the initial first layer neural network to obtain the predicted rotational speed. At the same time, the training battery pack temperature, the training ambient temperature, the training driving parameters, the training compressor speed, and the training electronic expansion valve opening are input into the initial second layer neural network to obtain the predicted temperature. This process continues until the predicted rotational speed meets the training compressor speed and the predicted temperature meets the training refrigerant temperature. If the predicted rotational speed satisfies the training compressor rotational speed and the predicted temperature satisfies the training refrigerant temperature, then the initial first layer neural network is determined to be the first layer neural network and the initial second layer neural network is determined to be the second layer neural network.

9. The method according to claim 8, wherein, Determining whether the predicted rotational speed meets the training compressor rotational speed and whether the predicted temperature meets the training refrigerant temperature includes: Obtain the first mean square error and the first average absolute percentage error between the predicted rotational speed and the training compressor rotational speed, and obtain the second mean square error and the second average absolute percentage error between the predicted temperature and the training refrigerant temperature; In response to the first mean square error being less than the first preset mean square error, and the first mean absolute percentage error being less than the first preset percentage error, and the second mean square error being less than the second preset mean square error, and the second mean absolute percentage error being less than the second preset percentage error, it is determined that the predicted speed satisfies the training compressor speed, and the predicted temperature satisfies the training refrigerant temperature. In response to the first mean square error being greater than or equal to the first preset mean square error, or the first average absolute percentage error being greater than or equal to the first preset percentage error, or the second mean square error being greater than or equal to the second preset mean square error, or the second average absolute percentage error being greater than or equal to the second preset percentage error, it is determined that the predicted speed does not meet the training compressor speed, or the predicted temperature does not meet the training refrigerant temperature.

10. The method according to claim 7, wherein, Before inputting the cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature into the initial first layer of the initial neural network model to obtain the predicted rotational speed, and simultaneously inputting the training battery pack temperature, the training ambient temperature, the training driving parameters, the training compressor speed, and the training electronic expansion valve opening into the initial second layer of the initial neural network model to obtain the predicted temperature, the method further includes: Identify anomalous data in the training dataset; Determine the average value of the two data points adjacent to the abnormal data; The abnormal data are replaced with the average value to obtain the processed training dataset.

11. The method according to claim 10, wherein, Identifying anomalous data in the training dataset includes: Obtain the average value of the training data in the training dataset; Based on the average value, determine the residuals and standard errors of the training data; At least one training data point whose residual is greater than a preset residual value, or whose standard error is greater than a preset standard error value, is identified as the abnormal data.

12. The method according to claim 7, wherein, Before inputting the cooling fan duty cycle, the training ambient temperature, and the training battery pack temperature into the initial first layer of the initial neural network model to obtain the predicted rotational speed, and simultaneously inputting the training battery pack temperature, the training ambient temperature, the training driving parameters, the training compressor speed, and the training electronic expansion valve opening into the initial second layer of the initial neural network model to obtain the predicted temperature, the method further includes: The training data in the training dataset is normalized to obtain a normalized training dataset.

13. A control device for a vehicle battery cooling system, comprising: The acquisition module is configured to acquire state parameters corresponding to the current state of the vehicle based on the current state of the vehicle. The state parameters include at least one of the following: the ambient temperature of the environment in which the vehicle is located, the driving parameters of the vehicle, the battery pack temperature, the opening degree of the electronic expansion valve, and the compressor speed. The battery pack, the electronic expansion valve, and the compressor are located in the battery cooling system of the vehicle. The processing module is configured to input the state parameters into a target neural network model to obtain target control parameters for the battery cooling system. The target control parameters include one of the following: the target rotational speed of the battery cooling system and the target temperature of the battery cooling system. The target rotational speed is used to characterize the predicted compressor rotational speed when the vehicle is in a non-driving state in the future time period. The target temperature is used to characterize the predicted refrigerant temperature at the battery cold plate outlet when the vehicle is in a driving state in the future time period. The battery cold plate outlet is located in the battery cooling system. The control module is configured to control the vehicle's battery cooling system based on the target control parameters.

14. An electronic device, comprising: Memory, which stores executable programs; A processor is configured to run the program, wherein the program executes the following method during runtime: Based on the current state of the vehicle, obtain state parameters corresponding to the current state, wherein the state parameters include at least one of the following: ambient temperature of the vehicle's environment, vehicle driving parameters, battery pack temperature, electronic expansion valve opening, and compressor speed, wherein the battery pack, electronic expansion valve, and compressor are located in the vehicle's battery cooling system; input the state parameters into a target neural network model to obtain target control parameters for the battery cooling system, wherein the target control parameters include one of the following: target speed of the battery cooling system and target temperature of the battery cooling system, wherein the target speed is used to characterize the predicted compressor speed when the vehicle is in a non-driving state in the future time period, and the target temperature is used to characterize the predicted refrigerant temperature at the battery cold plate outlet when the vehicle is in a driving state in the future time period, wherein the battery cold plate outlet is located in the battery cooling system; and control the vehicle's battery cooling system based on the target control parameters.

15. The electronic device according to claim 14, wherein, The program also executes the following methods during runtime: In response to the vehicle being in a non-driving state, the ambient temperature and the battery pack temperature are acquired to obtain the state parameters; In response to the vehicle being in a driving state, the ambient temperature, driving parameters, compressor speed, electronic expansion valve opening, and battery pack temperature are acquired to obtain the state parameters.

16. The electronic device according to claim 14, wherein, The target neural network model includes: a first-layer neural network and a second-layer neural network. The network structures of the first-layer neural network and the second-layer neural network are different. The program also executes the following methods during runtime: In response to the vehicle being in a non-driving state, the state parameters are input into the first layer of the neural network to obtain the target rotational speed of the battery cooling system; In response to the vehicle being in a driving state, the state parameters are input into the second layer of the neural network to obtain the target temperature of the battery cooling system.

17. The electronic device according to claim 14, wherein, The program also executes the following methods during runtime: In response to the vehicle being in a stationary state, the battery cooling system is controlled to operate based on the target rotational speed; In response to the vehicle being in motion, the opening of the electronic expansion valve of the battery cooling system or the speed of the compressor is adjusted based on the target temperature.

18. The electronic device according to claim 17, wherein, The program also executes the following methods during runtime: The pressure of the vehicle is obtained by a pressure sensor installed on the vehicle. Based on a preset correspondence, the refrigerant saturation temperature at the outlet of the battery cold plate corresponding to the pressure is determined; The difference between the target temperature and the refrigerant saturation temperature is obtained to determine the superheat. The opening degree of the electronic expansion valve of the battery cooling system or the compressor speed is adjusted based on the superheat.

19. A computer-readable storage medium comprising a stored executable program, wherein, When the executable program is executed, it controls the device containing the storage medium to perform the method described in any one of claims 1 to 12.

20. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 12.

Images