Vehicle heat management adjusting method, device, equipment and medium

By using the deep learning model of water pump speed and particle swarm algorithm in the automotive thermal management system to optimize the thermostat opening, the problem of traditional PID control response lag is solved, more efficient thermal management is achieved, and power output and ride comfort is improved. .

CN118997905APending Publication Date: 2024-11-22DONGFENG LIUZHOU MOTOR +1
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410951417.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-16
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

When traditional PID control faces the rapidly changing working conditions of modern automobile power systems, it has lagged response, increased energy consumption, and electronic water pumps lack efficient and precise control strategies, resulting in inefficient thermal management.

Method used

The deep learning model of water pump speed is used to analyze the vehicle state data, and the particle swarm algorithm and fuzzy controller optimization are used to adjust the thermostat opening to achieve accurate adjustment of the vehicle temperature.

Benefits of technology

It improves the response speed and efficiency of the vehicle thermal management system, reduces oscillation, shortens the temperature recovery time, and improves power output and ride comfort.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118997905A_ABST
    Figure CN118997905A_ABST
Patent Text Reader

Abstract

The invention discloses a vehicle thermal management adjusting method and device, equipment and a medium, and relates to the technical field of vehicle thermal management, and the method comprises the steps: obtaining vehicle state data, analyzing the vehicle state data through a water pump rotating speed deep learning model, obtaining an analysis result, obtaining the opening degree of a vehicle thermostat through the vehicle state data, and obtaining the opening degree of the vehicle thermostat; and the vehicle temperature is adjusted according to the analysis result and the opening degree of the vehicle thermostat. The rotating speed of the water pump is accurately predicted through deep learning, vehicle temperature optimization management is achieved by combining intelligent regulation and control of the state of the thermostat, and the heat management efficiency is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of vehicle thermal management technology, and in particular to a vehicle thermal management adjustment method, device, equipment and medium. Background Art

[0002] With the rapid development of the automotive industry and increasingly stringent environmental protection requirements, automotive thermal management technology has become one of the key areas for improving vehicle performance, energy conservation and emission reduction. While pursuing higher power density, modern automobile engines have put forward higher requirements for the efficiency and intelligence of thermal management systems. Traditional cooling systems, such as the configuration of mechanical water pumps and wax thermostats, have gradually failed to meet the complex and changeable working conditions of modern automobile power systems. Especially in new energy vehicles and hybrid vehicles, precise control of the thermal management system is particularly important because it directly affects the vehicle's range, battery life, power system efficiency and passenger comfort.

[0003] Traditional PID control has fixed parameters, which easily leads to large overshoot, long adjustment time, slow control process and other problems when disturbed, and has poor dynamic control performance. In contrast, the self-learning ability of neural network control can automatically adjust the control parameters to adapt to changes in working conditions. Through the adaptive effect on the system model, it can quickly respond to disturbances, greatly reduce the oscillation amplitude, shorten the transition time, and significantly shorten the time it takes for the controlled parameters to return to the set value.

[0004] Traditional control strategies, such as PID controllers, have fixed parameters and may not be able to adjust quickly when faced with rapidly changing operating conditions, resulting in delayed system response and increased energy consumption. Although electronic water pumps can be actively adjusted, they may operate in an inefficient area and increase energy consumption when there is a lack of efficient and accurate control strategies. Therefore, a smarter and more efficient thermal management control strategy is needed. Summary of the invention

[0005] The main purpose of this application is to provide a vehicle thermal management adjustment method, device, equipment and medium, aiming to solve the technical problem of how to improve the efficiency of vehicle thermal management adjustment.

[0006] To achieve the above objectives, the present application proposes a vehicle thermal management adjustment method, the method comprising:

[0007] Get vehicle status data;

[0008] Analyze the vehicle status data through a water pump speed deep learning model to obtain an analysis result;

[0009] Obtaining the vehicle thermostat opening through the vehicle status data;

[0010] The vehicle temperature is adjusted according to the analysis result and the vehicle thermostat opening.

[0011] Optionally, before the step of analyzing the vehicle status data by using the water pump speed deep learning model to obtain the analysis result, the step includes:

[0012] Establish an initial water pump speed deep learning model, and set an input layer, a hidden layer, and an output layer in the initial water pump speed deep learning model;

[0013] The initial water pump speed deep learning model is trained according to the vehicle state data to obtain a preset water pump speed deep learning model.

[0014] Optionally, the step of training the initial water pump speed deep learning model according to the vehicle state data to obtain a preset water pump speed deep learning model includes:

[0015] Initialize model parameters;

[0016] Get input data;

[0017] Inputting the input data into the initial water pump speed deep learning model for calculation to obtain predicted output data;

[0018] According to the loss function calculation, the error value between the predicted output data and the actual data is obtained;

[0019] By calculating through the back propagation algorithm, the gradient of the model parameters is obtained;

[0020] The model parameters are iteratively updated through an optimization algorithm according to the gradient until a maximum number of iterations is reached or the error value calculated by the loss function converges to a preset threshold, thereby obtaining a preset water pump speed deep learning model.

[0021] Optionally, after training the initial water pump speed deep learning model according to the vehicle state data to obtain a preset water pump speed deep learning model, the method further comprises:

[0022] Evaluate the preset water pump speed deep learning model to obtain an evaluation result;

[0023] If the evaluation result does not meet the thermal management control requirement, retraining the preset water pump speed deep learning model until the thermal management control requirement is met;

[0024] If the evaluation result meets the thermal management control requirements, the preset water pump speed deep learning model is used as the final water pump speed deep learning model, and the vehicle status data is analyzed.

[0025] Optionally, before the step of obtaining the vehicle thermostat opening through the vehicle status data, the following steps are included:

[0026] The membership function and the deep learning model of the pump speed are optimized by particle swarm algorithm;

[0027] The optimization of the membership function and the deep learning model of the water pump speed by using the particle swarm algorithm includes:

[0028] Initializing the positions and velocities of particles in the particle swarm and calculating the fitness of the particles;

[0029] Set the number of iterations and calculate the historical optimal value of the particle;

[0030] By comparing the fitness of the particles in each iteration, a global optimal value of the particle swarm is obtained;

[0031] According to the historical optimal value and the global optimal value, the position and speed of the particle are updated;

[0032] Determine whether the position and speed of the particle meet a preset range, and if not, perform boundary condition processing on the particle;

[0033] If the number of iterations is reached, the membership function and the water pump speed deep learning model are output;

[0034] The step of obtaining the vehicle thermostat opening through the vehicle status data comprises:

[0035] Constructing a fuzzy controller according to the membership function;

[0036] Obtaining a current coolant temperature according to the vehicle status data;

[0037] The vehicle thermostat opening is obtained through a fuzzy controller according to the current coolant temperature and the preset coolant temperature.

[0038] Optionally, the step of acquiring vehicle status data includes:

[0039] Get vehicle component parameters;

[0040] According to the vehicle component parameters, construct a whole vehicle cooling system simulation model;

[0041] The simulation model is used to perform simulation to obtain vehicle status data.

[0042] Optionally, the step of adjusting the vehicle temperature according to the analysis result and the vehicle thermostat opening comprises:

[0043] The predicted water pump speed is obtained through the analysis results;

[0044] adjusting the coolant working strategy according to the vehicle thermostat opening;

[0045] The vehicle temperature is adjusted according to the predicted water pump speed and the coolant operating strategy.

[0046] In addition, to achieve the above-mentioned purpose, the present application also proposes a vehicle thermal management adjustment device, the vehicle thermal management adjustment device comprising:

[0047] An acquisition module, used to acquire vehicle status data;

[0048] A processing module, used for analyzing the vehicle status data through a water pump speed deep learning model to obtain an analysis result;

[0049] An acquisition module, used to obtain the vehicle thermostat opening through the vehicle status data;

[0050] An execution module is used to adjust the vehicle temperature according to the analysis result and the vehicle thermostat opening.

[0051] In addition, to achieve the above-mentioned purpose, the present application also proposes a medium, which is a computer-readable medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the vehicle thermal management adjustment method as described above are implemented.

[0052] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the vehicle thermal management adjustment method described above are implemented.

[0053] This application obtains vehicle status data, analyzes the vehicle status data through a deep learning model of water pump speed, obtains analysis results, obtains the vehicle thermostat opening through the vehicle status data, and adjusts the vehicle temperature according to the analysis results and the vehicle thermostat opening. Use deep learning to accurately predict the water pump speed, combined with intelligent regulation of the thermostat state, to achieve optimized vehicle temperature management and improve thermal management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0055] Figure 1 This is a flow chart of the first embodiment of the vehicle thermal management adjustment method of the present application;

[0056] Figure 2 A schematic diagram of a water pump control process of the first embodiment of the vehicle thermal management adjustment method of the present application;

[0057] Figure 3 This is a flow chart of a second embodiment of the vehicle thermal management adjustment method of the present application;

[0058] Figure 4 This is a schematic diagram of the module structure of the vehicle thermal management adjustment device according to an embodiment of the present application;

[0059] Figure 5 Schematic diagram of the device structure of the hardware operating environment involved in the vehicle thermal management adjustment method in the embodiment of the present application.

[0060] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0061] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0062] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0063] The main solution of the embodiment of the present application is: by acquiring vehicle status data, analyzing the vehicle status data through a water pump speed deep learning model to obtain analysis results, obtaining the vehicle thermostat opening through the vehicle status data, and adjusting the vehicle temperature according to the analysis results and the vehicle thermostat opening.

[0064] Based on this, the embodiment of the present application provides a vehicle thermal management adjustment method, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the vehicle thermal management adjustment method of the present application.

[0065] In this embodiment, the vehicle thermal management adjustment method includes steps S10 to S40:

[0066] Step S10, obtaining vehicle status data.

[0067] It should be noted that before obtaining the vehicle status data, the vehicle component parameters are obtained, including environmental parameters, vehicle weight, driving resistance, speed ratio, transmission parameters, engine parameters, cooling system (pipeline, electronic water pump, radiator, thermostat) and other parameters. According to the vehicle component parameters, a simulation model of the vehicle cooling system is constructed, and the simulation model is used to perform simulation to obtain the vehicle status data. After the model is established, a simulation test is performed under a cyclic working condition to obtain the engine state of the vehicle under the driving condition and the state parameter data of each component of the cooling system. According to the simulation results, the bench test is compared, and the key parameters such as the HVAC outlet water temperature, the supercharger outlet water temperature, the HVAC inlet water flow rate and the oil coolant temperature are mainly checked to ensure that the error is within the allowable range.

[0068] Step S20: Analyze the vehicle status data through the water pump speed deep learning model to obtain analysis results.

[0069] It should be noted that the water pump speed deep learning model is established based on the neural network model, including but not limited to the use of the BP neural network model. The model first integrates the multi-sensor data inside and outside the vehicle, including key parameters such as the temperature and humidity in the cab, the ambient temperature outside the vehicle, the battery temperature, and the working status of the radiator. After preprocessing, these data are used as the input of the deep learning model. The model uses the BP neural network architecture to perform deep learning and feature extraction on the input data through a series of linear transformations and nonlinear activation functions, such as Tanh or Linear functions. During the training process, the model uses a large amount of historical data to optimize the weights and bias parameters through the back propagation algorithm to minimize the error between the predicted output and the actual label. The mean square error (MSE) or cross-entropy loss (Cross-Entropy Loss) is often used as the loss function. This process not only enables the model to learn the complex patterns of changes in the environment and vehicle status, but also accurately predicts the ideal working state of each component of the thermal management system in a new environment.

[0070] Step S30, obtaining the vehicle thermostat opening through the vehicle status data.

[0071] It is important to understand that vehicle status data covers a wide range of information, such as engine coolant temperature, ambient temperature, vehicle speed, engine load, air conditioning system operating status, etc. These data are transmitted to the vehicle's electronic control unit (ECU) in real time through the vehicle network. The ECU is embedded with complex algorithms, which may include traditional control logic based on physical models, fuzzy logic control, and even machine learning models, which work together on these raw data for comprehensive analysis and processing.

[0072] For the decision of thermostat opening, the system first compares the coolant temperature with the preset ideal operating temperature range to determine the current cooling demand. Then, combined with dynamic information such as engine load and vehicle speed, the coolant temperature change trend in the future is predicted. For example, when driving at high speed or under heavy load, even if the current coolant temperature is moderate, the system will predict the risk of overheating in the future, so it will increase the thermostat opening in advance to enhance the heat dissipation capacity. In order to obtain the vehicle thermostat opening, the membership function and the preset water pump speed deep learning model need to be optimized through the particle swarm algorithm.

[0073] The membership function and the deep learning model of the water pump speed are further optimized by the particle swarm algorithm. Taking the optimization of the deep learning model of the water pump speed by the particle swarm algorithm as an example, the position and speed of the particles in the particle swarm are initialized and the fitness of the particles is calculated. The number of iterations is set, and the historical optimal value of the particles is calculated. By comparing the fitness of the particles in each iteration, the global optimal value of the particle swarm is obtained; according to the historical optimal value and the global optimal value, the position and speed of the particles are updated. The specific update formula is:

[0074] v in (t+1)=wv in (t)+r1c1(p in (t)-x in (t))+r2c2(p bin (t)-x in (t))

[0075] x in (t+1)=x in (t)+v in (t+1)

[0076] Among them, w is the inertia weight of the particle swarm; r1, r2 are random constants ranging from 0 to 1; c1, c2 are learning factors; v in is the nth dimension of the velocity of the ith particle; x in (t) is the nth dimension of the position of the ith particle; p in is the nth dimension of the optimal position of the particle; p bin is the nth dimension of the position of the entire population. The speed update formula is divided into three parts. The first part is the inheritance of the current speed state, which is related to the inertia weight. The second part represents the learning of the best position for itself, and the degree of learning is related to the learning factor. The third part represents the learning of the optimal position of the population, which is also related to the learning factor. This part will guide the particles to move to the global optimal position.

[0077] When the algorithm starts to operate, the inertia weight is as large as possible, so that the algorithm can reach the global optimum faster. Therefore, the inertia weight at the beginning of the algorithm is 0.9, and the inertia weight at the convergence moment is small so as not to miss the global optimum. The value at the convergence moment is 0.4. The learning factors and are the learning of individual particles and populations. In the initial operation stage of the algorithm, the learning of individual particles is very important for obtaining the global optimal solution, and the learning of the population in the later stage of the algorithm operation is the key to obtaining the optimal solution. Therefore, in this embodiment, the learning factors and are set to 2. Determine whether the position and speed of the particle meet the preset range. If not, the particle is subjected to boundary condition processing. If the number of iterations is reached, the membership function and the deep learning model of the pump speed are output. In this embodiment, the maximum number of iterations is set to 100.

[0078] As for the optimization of the membership function of the fuzzy controller constructed by the particle swarm algorithm, the above steps are the same. Since the particle swarm algorithm optimizes the membership function model of the fuzzy controller, and the fuzzy controller has two inputs and one output, the input fuzzy language parameters are five in total, and the triangular membership function used is symmetrical, so one input needs to optimize two variables, two inputs need to optimize four, and the output fuzzy language parameters are seven in total. Similarly, there are three parameters that need to be optimized, and the input and output parameters that need to be optimized are seven in total. The optimization range of the particle swarm algorithm is plus or minus 0.5 on the basis of the original value, which constitutes the variable range of the parameters to be optimized. The seven parameters to be optimized are used as the input of the particle swarm algorithm. The value range of the particle swarm algorithm input e and ec is [-3, 3], and a value is taken every 0.1, which fully covers the value range of the input.

[0079] Among them, the fitness function is set and the expression is:

[0080]

[0081] In the formula: a, b and c are weighted parameters of input and output, representing the proportion of input and output in the fitness function. In the present invention, a, b and c are 0.25, 0.35 and 0.4 respectively; e is the coolant temperature difference; ec is the temperature difference change rate; u represents the change in the thermostat opening. This formula can cover all input values. The larger the denominator in this formula, the smaller its fitness, which means that the thermostat will oscillate less and more gently.

[0082] Furthermore, the step of obtaining the vehicle thermostat opening is performed through the vehicle state data. First, a fuzzy controller is constructed according to the membership function. The fuzzy controller of the vehicle thermostat opening selects the temperature difference e between the coolant temperature and the target temperature within a sampling period and the temperature difference change rate ec as input. The domain of the input variable is [-3, 3]. The output is the opening rate of the thermostat. The domain of the output is [-1, 1]. The fuzzy subset describes the size of the input variable using human language. The size of the input variable is distinguished by dividing the domain of the input variable into multiple order levels. In this embodiment, the fuzzy subset of the input variable is set to {NB (negative large), NS (negative small), ZO (zero), PS (positive large), PB (positive large)}, and the fuzzy subset of the output variable is set to {NB (negative large), NM (negative medium), NS (negative small), ZO (zero), PS (positive small), PM (positive medium), PB (positive large)}. The membership function is used to determine the membership of the input or output precise value to the fuzzy language value in a fuzzy subset, and is used to describe the degree to which the input or output definite value belongs to a fuzzy language. In this embodiment, the membership is a triangular membership function, and the specific expression is:

[0083]

[0084] Among them, b is the horizontal coordinate of the vertex of the triangular membership function, and a and c are the horizontal coordinates of the base.

[0085] The fuzzy rules of the fuzzy controller are some conditional statements formed by combining the fuzzy language in the fuzzy subset. The fuzzy rules form a relationship table of the relationship between the thermostat temperature difference and the temperature difference change rate. When performing fuzzy reasoning, the fuzzy quantity can be converted into an accurate value according to the fuzzy rules. Using if-then fuzzy control rules, Table 1 is a fuzzy control rule table. For example, NB in ​​E means negative large, which means that the input variable E is much smaller than the target value, indicating that the cooling water temperature is relatively low. The NB in ​​the EC temperature difference change rate means that the temperature drops relatively quickly, and the output NB means that the thermostat opening should be reduced a lot. If E=NS and EC=ZO then U=NM, it means that when the cooling water temperature is low and the temperature drops relatively quickly, the thermostat opening should be reduced a little.

[0086] Table 1 Fuzzy control rules table

[0087]

[0088]

[0089] The fuzzy reasoning function Mamdani method is used to use fuzzy rules to infer the output state of the fuzzy value. Finally, the centroid method is used to defuzzify and transform the fuzzy value after fuzzy reasoning into an accurate value. The centroid method selects the variable value corresponding to the centroid of the graph synthesized by the function curve and the basic variable axis as the output value.

[0090] The current coolant temperature is obtained according to the vehicle status data, and the vehicle thermostat opening is obtained through the fuzzy controller according to the current coolant temperature and the preset coolant temperature. After receiving the coolant temperature data, the ECU performs necessary preprocessing, such as filtering, calibration, etc., to eliminate noise and improve the credibility of the data. Subsequently, the current coolant temperature is compared with the preset coolant temperature to calculate the temperature difference and its rate of change. The temperature difference and its rate of change are converted into fuzzy language variables, such as "large", "medium", "small" or "very low", "low", "medium", "high", "very high". This step is completed by defining appropriate membership functions, which map specific values ​​to the degree in the fuzzy set. According to the preset fuzzy control rule table, the above fuzzified input (temperature difference and rate of change) is matched with the rules. The fuzzy rule base contains a series of condition-action rules established based on expert experience or historical data, such as "If the temperature difference is large and the rate of change is fast, the thermostat opening should be large". Fuzzy logic operations combined with fuzzy rules are used to reason the input fuzzy set to obtain the fuzzy output set of the thermostat opening. This output set reflects the degree of fuzziness of how the thermostat should be adjusted in the current state. Finally, the fuzzy output is converted into a specific thermostat opening value through defuzzification techniques (such as the centroid method and the maximum membership method). The defuzzification process determines the clear value that best represents the fuzzy output set, that is, the actual opening instruction of the thermostat. According to the defuzzified opening value, the ECU sends a control signal to the electronic thermostat to adjust the opening and closing degree of the thermostat valve.

[0091] Step S40, adjusting the vehicle temperature according to the analysis result and the vehicle thermostat opening.

[0092] It should be noted that the predicted water pump speed is obtained through the analysis results, and the coolant working strategy is adjusted according to the vehicle thermostat opening. The vehicle temperature is adjusted according to the predicted water pump speed and coolant working strategy. This adjustment process combines advanced algorithm analysis and real-time control technology to ensure that the engine can maintain the optimal operating temperature under various working conditions, thereby optimizing fuel economy, improving power output, reducing emissions, and enhancing vehicle durability and passenger comfort.

[0093] like Figure 2As shown in the schematic diagram of the water pump control process, first, the water pump speed deep learning model optimized by the particle swarm algorithm is stored in the engine ECU, and the sensors used to detect the real-time status of the vehicle and the PWM controller are connected to the ECU through the CAN bus. The detected vehicle speed, BMEP, engine speed, coolant temperature and thermostat opening are used as model inputs. The water pump speed is predicted by the neural network prediction model according to the real-time data. The PWM receives the command instructions from the ECU, and uses the predicted speed as the basis for the change of the input duty cycle, and then controls the motor voltage according to the duty cycle of the pulse signal. The motor controls the water pump speed according to the voltage, so that it runs at the lowest power consumption speed under the heat dissipation conditions, and instructs the electronic water pump to increase the speed, appropriately increase the thermostat opening, promote more coolant to pass through the radiator, and accelerate heat exchange. On the contrary, if the prediction shows that the current cooling is excessive, the system will slow down the water pump speed and reduce the thermostat opening to reduce heat dissipation and keep the engine warm, especially in cold weather or at the beginning of vehicle startup.

[0094] This embodiment provides a vehicle thermal management adjustment method, which obtains vehicle status data, analyzes the vehicle status data through a water pump speed deep learning model, obtains analysis results, obtains the vehicle thermostat opening through the vehicle status data, and adjusts the vehicle temperature according to the analysis results and the vehicle thermostat opening. The water pump speed is accurately predicted by deep learning, combined with intelligent regulation of the thermostat state, to achieve vehicle temperature optimization management and improve thermal management efficiency.

[0095] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can refer to the above introduction, and will not be repeated later. Figure 3 The vehicle thermal management adjustment method further includes steps S201 to S202 before step S20:

[0096] Step S201, establish an initial water pump speed deep learning model, and set an input layer, a hidden layer and an output layer in the initial water pump speed deep learning model.

[0097] It should be noted that the initial water pump speed deep learning model is established, and the parameters of the initial water pump speed deep learning model are updated and improved through machine learning methods, so as to manage the vehicle temperature intelligently, ensure that the vehicle thermal management system can operate efficiently under various working conditions, and improve energy efficiency and ride comfort. The core design of the model revolves around the input layer, hidden layer and output layer. Each layer plays a key role and together constitutes the neural network architecture of the model.

[0098] The input layer is the starting point of the model. The number of nodes in the input layer is the same as the dimension of the data input. Five parameters, namely vehicle speed, BMEP, engine speed, coolant temperature and thermostat opening, are selected as input data, which means that the number of input layer nodes is 5. The number of nodes in the input layer matches the number of features collected, ensuring that the model can fully capture the environment and vehicle status information related to thermal management.

[0099] The hidden layer is responsible for extracting and transforming complex features of the input data. Usually, multiple hidden layers are designed, and each layer uses a different activation function (such as ReLU, tanh, etc.) to introduce nonlinear relationships so that the model can learn complex patterns and associations in the data. The number of nodes and layers in the hidden layer is determined through repeated experiments, aiming to balance the complexity of the model with the risk of overfitting, ensuring that the model has good learning ability and does not overfit the training data. The specific empirical formula for node calculation is:

[0100]

[0101] Among them, m is the number of hidden layer nodes, n is the number of input layer nodes, and 1 represents the number of output layer nodes. According to the setting of 5 input layer nodes and 1 output layer nodes, a is usually a constant between 1 and 10, and the value range of m is between 3 and 13. By controlling other variables except the number of hidden layer nodes, the number of nodes is adjusted between 3 and 13, and the number of hidden layer nodes that need to be changed is brought into the neural network model from small to large. The overall error R value of the degree of fit between the output value and the target value of the network model is calculated. The closer the R value is to 1, the higher the degree of fit between the output value and the target value, and the better the prediction effect. When the number of neural network nodes is 11, the neural training is best, and the output value and the target value are better.

[0102] The output layer determines the prediction or decision information that the model ultimately provides. In a thermal management system, the output layer may include recommended values ​​for control parameters such as fan speed, water pump flow rate, and coolant flow rate. The number of nodes in the output layer should be consistent with the dimension of the thermal management parameter that needs to be controlled. Because the output layer of the BP neural network is the electronic water pump speed, the number of nodes in the output layer is 1. Through the deep learning process, the model will learn how to output the most optimized control instructions based on the input information to achieve precise adjustment of the vehicle thermal management system.

[0103] Step S202: training the initial water pump speed deep learning model according to the vehicle state data to obtain a preset water pump speed deep learning model.

[0104] It should be understood that to train the initial water pump speed deep learning model, first initialize the model parameters, obtain the input data, and then input the input data into the initial water pump speed deep learning model to calculate and obtain the predicted output data. Then, according to the loss function calculation, the error value between the predicted output data and the actual data is obtained, and the gradient of the model parameters is calculated through the back propagation algorithm. Finally, according to the gradient, the model parameters are iteratively updated through the optimization algorithm until the maximum number of iterations is reached or the error value calculated by the loss function converges to the preset threshold, and the preset water pump speed deep learning model is obtained. Specifically, first, vehicle status data such as the engine status and status parameter data of each component of the cooling system under driving conditions are collected, cleaned, and z-score normalized. The original data is subtracted from the data mean and then divided by the data standard deviation. The data after normalization will obey the distribution with a mean of 0 and a variance of 1. The formula is:

[0105]

[0106] According to the above formula, noise and outliers are eliminated to form a high-quality training data set.

[0107] The data set is divided into training set and test set in a certain ratio, generally with a ratio of 70% and 30%, to ensure that the training, tuning and generalization ability evaluation of the model are scientific and reasonable. Then, these data sets are used to train the initial model. The model adopts a multi-layer perceptron structure, including a carefully designed input layer, multiple hidden layers and an output layer. During the training process, the model predicts the output through forward propagation calculation, and the input data of the input layer is calculated through linear change to obtain the input data of the hidden layer. Specifically, the input data of the hidden layer is obtained by the following formula:

[0108] z( 1 )=W( 1 )x+b( 1 )

[0109] where z (1) is the input value of the first hidden layer, W (1) is the weight matrix of the first hidden layer, which indicates the connection strength between the input layer and the hidden layer. x is the input data of the input layer, and b (1) It is a bias vector used to adjust the activation value of the hidden layer. The superscript represents the data of the hidden layer. Each hidden layer has a different weight matrix and bias vector.

[0110] The input data of the hidden layer is calculated through a nonlinear activation function to obtain the output data of the hidden layer. Specifically, it is obtained by the following formula:

[0111] a (1) =Tanh(z (1) )

[0112]

[0113] Among them, a (1) The output data of the first hidden layer. According to the above two formulas, the output data of the hidden layer is first subjected to linear change calculation and then to nonlinear activation function calculation until the output data of the last hidden layer is obtained. The output data of the last hidden layer is used as the input data of the output layer, and is calculated according to the nonlinear activation function to obtain the predicted output data. Specifically, it is obtained according to the following formula:

[0114] Z( 3 )=W( 3 )A( 2 )+b( 3 )

[0115] Y=Linear(Z (3) )

[0116] Linear(x)=x

[0117] Among them, Z (3) is the input value of the output layer, W (3) is the weight matrix of the output layer, which represents the connection strength between the output layer and the second hidden layer. (2) is the input data of the second hidden layer, b (3) is the bias vector, and Y is the output value of the output layer.

[0118] The error value between the predicted result and the actual result is obtained through the loss function (such as mean square error MSE or cross entropy loss). Then, the gradient of the loss function with respect to the model parameters is calculated using the back propagation algorithm. Specifically, the error gradient of the output layer is obtained according to the error value. Specifically, it is obtained by the following formula:

[0119]

[0120] Where L is the loss function and σ is the derivative of the output layer activation function.

[0121] According to the chain rule, the error gradient of the output layer is combined with the derivative of the activation function of the previous layer to calculate the error gradient of the previous layer. Specifically, it is obtained by the following formula:

[0122]

[0123] According to the above formula, the hidden layer before the output layer is obtained, and then the error gradient of other layers of the deep learning model is obtained according to the above formula, and the gradient of the model parameters of each layer of the initial water pump speed deep learning model is obtained, including the gradient of the weight matrix Gradient of the bias vector

[0124] Then, the stochastic gradient descent optimization strategy is adopted to obtain the learning rate η and gradually adjust the model parameters until the loss function converges or reaches the predetermined number of iterations. This process is repeated to gradually approach the global optimal solution. During the entire training phase, it is also necessary to pay close attention to the overfitting problem of the model and use regularization, dropout and other technologies to alleviate it when appropriate. The model hyperparameters, such as learning rate and number of hidden layer nodes, are continuously adjusted through cross-validation and other means to optimize the model performance. After obtaining the preset water pump speed deep learning model, the preset water pump speed deep learning model is evaluated to obtain the evaluation result. If the evaluation result does not meet the thermal management control requirements, the preset water pump speed deep learning model is retrained until it meets the thermal management control requirements; if the evaluation result meets the thermal management control requirements, the preset water pump speed deep learning model is used as the final water pump speed deep learning model, and the vehicle status data is analyzed.

[0125] The present embodiment establishes an initial water pump speed deep learning model, and sets an input layer, a hidden layer, and an output layer in the initial water pump speed deep learning model. The initial water pump speed deep learning model is trained according to the vehicle status data to obtain a preset water pump speed deep learning model. By establishing the water pump speed deep learning model, the water pump speed of the vehicle is controlled through training optimization, thereby regulating the thermal control system, and corresponding changes can be made in real time according to environmental changes.

[0126] This application also provides a vehicle thermal management adjustment device, please refer to Figure 4 , the device comprises:

[0127] The acquisition module 10 is used to acquire vehicle status data.

[0128] The processing module 20 is used to analyze the vehicle status data through the water pump speed deep learning model to obtain the analysis result.

[0129] The acquisition module 10 is used to obtain the vehicle thermostat opening through the vehicle status data.

[0130] The execution module 30 is used to adjust the vehicle temperature according to the analysis result and the vehicle thermostat opening.

[0131] The vehicle thermal management adjustment device provided by the present application adopts the vehicle thermal management adjustment method in the above embodiment, which can solve the technical problem of how to improve the efficiency of vehicle thermal management adjustment. Compared with the prior art, the beneficial effects of the vehicle thermal management adjustment device provided by the present application are the same as the beneficial effects of the vehicle thermal management adjustment method provided by the above embodiment, and the other technical features in the vehicle thermal management adjustment device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0132] In one embodiment, the processing module 20 is also used to establish an initial water pump speed deep learning model, and set an input layer, a hidden layer and an output layer in the initial thermal management deep learning model; the initial water pump speed deep learning model is trained according to the vehicle status data to obtain a preset water pump speed deep learning model.

[0133] In one embodiment, the processing module 20 is also used to initialize model parameters; obtain input data; input the input data into the initial water pump speed deep learning model to obtain predicted output data; calculate according to the loss function to obtain the error value between the predicted output data and the actual data; calculate through the back propagation algorithm to obtain the gradient of the model parameters; iteratively update the model parameters through the optimization algorithm according to the gradient until the maximum number of iterations is reached or the error value calculated by the loss function converges to a preset threshold, thereby obtaining a preset water pump speed deep learning model.

[0134] In one embodiment, the processing module 20 is also used to evaluate the preset water pump speed deep learning model to obtain an evaluation result; if the evaluation result does not meet the thermal management control requirements, the preset water pump speed deep learning model is retrained until the thermal management control requirements are met; if the evaluation result meets the thermal management control requirements, the preset water pump speed deep learning model is used as the final water pump speed deep learning model, and the vehicle status data is analyzed.

[0135] In one embodiment, the acquisition module 10 is also used to optimize the membership function and the preset water pump speed deep learning model through a particle swarm algorithm; the optimization of the membership function and the preset water pump speed deep learning model through a particle swarm algorithm includes: initializing the position and speed of particles in the particle swarm and calculating the fitness of the particles; setting the number of iterations and calculating the historical optimal value of the particles; obtaining the global optimal value of the particle swarm by comparing the fitness of the particles in each iteration; updating the position and speed of the particles according to the historical optimal value and the global optimal value; judging whether the position and speed of the particles meet the preset range, if not, performing boundary condition processing on the particles; if the number of iterations is reached, outputting the membership function and the preset water pump speed deep learning model; the step of obtaining the vehicle thermostat opening through the vehicle status data includes: constructing a fuzzy controller according to the membership function; obtaining the current coolant temperature according to the vehicle status data; obtaining the vehicle thermostat opening through the fuzzy controller according to the current coolant temperature and the preset coolant temperature.

[0136] In one embodiment, the acquisition module 10 is further used to acquire vehicle component parameters; construct a whole vehicle cooling system simulation model according to the vehicle component parameters; and obtain vehicle status data by simulating the simulation model.

[0137] In one embodiment, the execution module 30 is also used to obtain a predicted water pump speed through analysis results; adjust the coolant working strategy according to the vehicle thermostat opening; and adjust the vehicle temperature according to the predicted water pump speed and the coolant working strategy.

[0138] The present application provides a vehicle thermal management and adjustment device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the vehicle thermal management and adjustment method in the above-mentioned embodiment one.

[0139] Reference below Figure 5 , which shows a schematic diagram of the structure of a vehicle thermal management and adjustment device suitable for implementing the embodiment of the present application. The vehicle thermal management and adjustment device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The vehicle thermal management adjustment device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0140] like Figure 5As shown, the vehicle thermal management and adjustment device may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the vehicle thermal management and adjustment device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the vehicle thermal management and regulation device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a vehicle thermal management and regulation device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0141] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0142] The vehicle thermal management and adjustment device provided by the present application adopts the vehicle thermal management and adjustment method in the above embodiment, which can solve the technical problem of how to improve the efficiency of vehicle thermal management and adjustment. Compared with the prior art, the beneficial effects of the vehicle thermal management and adjustment device provided by the present application are the same as the beneficial effects of the vehicle thermal management and adjustment method provided by the above embodiment, and the other technical features in the vehicle thermal management and adjustment device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0143] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0144] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0145] The present application provides a computer-readable medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the vehicle thermal management adjustment method in the above-mentioned embodiment.

[0146] The computer-readable medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0147] The computer-readable medium may be included in the vehicle thermal management and regulation device; or may exist independently without being assembled into the vehicle thermal management and regulation device.

[0148] The above-mentioned computer-readable medium carries one or more programs. When the above-mentioned one or more programs are executed by the vehicle thermal management and adjustment device, the vehicle thermal management and adjustment device can write computer program codes for performing the operations of the present application in one or more programming languages ​​or a combination thereof. The above-mentioned programming languages ​​include object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0149] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0150] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0151] The readable medium provided by the present application is a computer-readable medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned vehicle thermal management adjustment method, and can solve the technical problem of how to improve the efficiency of vehicle thermal management adjustment. Compared with the prior art, the beneficial effects of the computer-readable medium provided by the present application are the same as the beneficial effects of the vehicle thermal management adjustment method provided by the above-mentioned embodiment, and will not be repeated here.

[0152] The present application also provides a computer program product, including a computer program, which implements the steps of the vehicle thermal management adjustment method as described above when executed by a processor.

[0153] The computer program product provided by this application can solve the technical problem of how to improve the efficiency of vehicle thermal management regulation. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the vehicle thermal management regulation method provided by the above embodiment, which will not be repeated here.

[0154] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A vehicle thermal management adjustment method, characterized in that: The method comprises: Get vehicle status data; Analyze the vehicle status data through a water pump speed deep learning model to obtain an analysis result; Obtaining the vehicle thermostat opening through the vehicle status data; The vehicle temperature is adjusted according to the analysis result and the vehicle thermostat opening.

2. The method according to claim 1, characterized in that Before the step of analyzing the vehicle status data by the water pump speed deep learning model to obtain the analysis result, the method includes: Establish an initial water pump speed deep learning model, and set an input layer, a hidden layer, and an output layer in the initial water pump speed deep learning model; The initial water pump speed deep learning model is trained according to the vehicle state data to obtain a preset water pump speed deep learning model.

3. The method according to claim 1, characterized in that The step of training the initial water pump speed deep learning model according to the vehicle state data to obtain a preset water pump speed deep learning model includes: Initialize model parameters; Get input data; Inputting the input data into the initial water pump speed deep learning model for calculation to obtain predicted output data; According to the loss function calculation, the error value between the predicted output data and the actual data is obtained; By calculating through the back propagation algorithm, the gradient of the model parameters is obtained; The model parameters are iteratively updated through an optimization algorithm according to the gradient until a maximum number of iterations is reached or the error value calculated by the loss function converges to a preset threshold, thereby obtaining a preset water pump speed deep learning model.

4. The method according to claim 1, characterized in that After the initial water pump speed deep learning model is trained according to the vehicle state data to obtain a preset water pump speed deep learning model, the method includes: Evaluate the preset water pump speed deep learning model to obtain an evaluation result; If the evaluation result does not meet the thermal management control requirement, retraining the preset water pump speed deep learning model until the thermal management control requirement is met; If the evaluation result meets the thermal management control requirements, the preset water pump speed deep learning model is used as the final water pump speed deep learning model, and the vehicle status data is analyzed.

5. The method according to claim 1, characterized in that Before the step of obtaining the vehicle thermostat opening through the vehicle status data, the method includes: The membership function and the preset pump speed deep learning model are optimized through particle swarm algorithm; The optimization of the membership function and the preset water pump speed deep learning model by using the particle swarm algorithm includes: Initializing the positions and velocities of particles in the particle swarm and calculating the fitness of the particles; Set the number of iterations and calculate the historical optimal value of the particle; By comparing the fitness of the particles in each iteration, a global optimal value of the particle swarm is obtained; According to the historical optimal value and the global optimal value, the position and speed of the particle are updated; Determine whether the position and speed of the particle meet a preset range, and if not, perform boundary condition processing on the particle; If the number of iterations is reached, the membership function and the preset water pump speed deep learning model are output; The step of obtaining the vehicle thermostat opening through the vehicle status data comprises: Constructing a fuzzy controller according to the membership function; Obtaining a current coolant temperature according to the vehicle status data; The vehicle thermostat opening is obtained through a fuzzy controller according to the current coolant temperature and the preset coolant temperature.

6. The method according to claim 1, characterized in that The step of obtaining vehicle status data comprises: Get vehicle component parameters; According to the vehicle component parameters, construct a whole vehicle cooling system simulation model; The simulation model is used to perform simulation to obtain vehicle status data.

7. The method according to claim 1, characterized in that The step of adjusting the vehicle temperature according to the analysis result and the vehicle thermostat opening comprises: The predicted water pump speed is obtained through the analysis results; adjusting the coolant working strategy according to the vehicle thermostat opening; The vehicle temperature is adjusted according to the predicted water pump speed and the coolant operating strategy.

8. A vehicle thermal management adjustment device, characterized in that: The device comprises: An acquisition module, used to acquire vehicle status data; A processing module, used for analyzing the vehicle status data through a water pump speed deep learning model to obtain an analysis result; An acquisition module, used to obtain the vehicle thermostat opening through the vehicle status data; An execution module is used to adjust the vehicle temperature according to the analysis result and the vehicle thermostat opening.

9. A vehicle thermal management and adjustment device, characterized in that: The device comprises: a memory, a processor, and a vehicle thermal management adjustment program stored in the memory and executable on the processor, wherein the vehicle thermal management adjustment program is configured to implement the steps of the vehicle thermal management adjustment method according to any one of claims 1 to 7.

10. A medium, characterized in that The medium stores a vehicle thermal management adjustment program, which, when executed by a processor, implements the steps of the vehicle thermal management adjustment method according to any one of claims 1 to 7.

Citation Information

Cited By

  • Dynamic control method of battery thermal management system based on deep learning

    CN120073150A