Motor temperature model construction method and device, electronic equipment and storage medium

Abstract

The invention provides a motor temperature model construction method and device, electronic equipment and a storage medium, and relates to the technical field of motors. The motor temperature model construction method comprises the steps of obtaining simulation data of a motor based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model; a preset model is trained according to the simulation data, a temperature prediction model is established, and the temperature prediction model is used for predicting the motor temperature. The simulation data of the motor are acquired based on the one-dimensional thermal network model and the three-dimensional fluid thermal coupling model, then the temperature prediction model is established based on the one-dimensional and three-dimensional simulation data, and the prediction model is driven by using different levels of data, so that a large number of bench tests and tedious three-dimensional simulation calculation can be avoided, quick response and high precision are ensured, and the method is suitable for large-scale popularization and application. The design period is remarkably shortened, the cost is reduced, and the temperature of the key component of the motor can be quickly and accurately predicted under complex working conditions.

Description

Technical Field

[0001] This invention relates to the field of motor technology, and more specifically, to a method, apparatus, electronic device, and storage medium for constructing a motor temperature model. Background Technology

[0002] As motor power continues to increase, the problem of internal heat generation becomes increasingly prominent. During motor operation, losses in components such as stator copper losses are converted into heat. If heat dissipation is not timely, the temperature of core components will rise, leading to decreased motor performance, shortened lifespan, and even safety risks. Therefore, motor thermal management and temperature prediction have become key aspects in motor design and application.

[0003] In related technologies, temperature data of the motor can be collected through bench tests, and then a temperature prediction model can be established using neural networks. However, the initial test data collection cycle is long and the cost is high. Alternatively, a thermal network model can be used to abstract the motor structure into several heat dissipation nodes, using heat capacity to characterize the thermal inertia of the components and thermal resistance to represent the conduction and convection characteristics. Combined with the power loss input, the node thermal balance equation can be solved to calculate the motor temperature. However, this is time-consuming for complex operating conditions and may result in serious deviations between the simulation results and the actual situation. Summary of the Invention

[0004] The problem solved by this invention is how to quickly and accurately predict motor temperature.

[0005] To address the above problems, this invention provides a method, apparatus, electronic device, and storage medium for constructing a motor temperature model.

[0006] In a first aspect, the present invention provides a method for constructing a motor temperature model, comprising: Simulation data of the motor were obtained based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model. The preset model is trained based on the simulation data to establish a temperature prediction model, which is used to predict the motor temperature.

[0007] Optionally, the acquisition of motor simulation data based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model includes: The one-dimensional thermal network model is constructed, and the component losses under different operating conditions are input into the one-dimensional thermal network model for simulation to obtain the first temperature data of multiple locations in the motor. The three-dimensional fluid thermal coupling model is constructed, and the component losses under different operating conditions are input into the three-dimensional fluid thermal coupling model for simulation to obtain second temperature data at multiple locations in the motor. The simulation data includes the first temperature data and the second temperature data.

[0008] Optionally, constructing the one-dimensional heat network model includes: The motor is divided into multiple hot nodes; The thermal capacity and thermal resistance are set according to the component materials and structural parameters of the motor, and the conduction and convection relationships between the thermal nodes are established. The various losses of the motor are used as heat sources input to the corresponding thermal nodes to construct the one-dimensional thermal network model.

[0009] Optionally, constructing the three-dimensional fluid thermal coupling model includes: A geometric model of the stator and windings of the motor is established, and the geometric model is discretized to form a mesh model; The temperature field and flow field are established based on the mesh model to construct the three-dimensional fluid thermal coupling model.

[0010] Optionally, training the preset model based on the simulation data to establish a temperature prediction model includes: The preset model is trained based on the first temperature data to establish a one-dimensional simulation-driven temperature prediction model. The preset model is trained based on the second temperature data to establish a three-dimensional simulation-driven temperature prediction model.

[0011] Optionally, training the preset model based on the first temperature data includes: A first data matrix is ​​constructed based on the first temperature data. The first data matrix includes a first input matrix and a first output matrix. The first input matrix includes motor component losses, motor speed, coolant flow rate, coolant temperature, and the temperature of each motor component. The first output matrix includes the temperature change rate of each motor component and the temperature change rate of the coolant. The first data matrix is ​​used as training data to train the preset model.

[0012] Optionally, training the preset model based on the second temperature data includes: A second data matrix is ​​constructed based on the second temperature data. The second data matrix includes a second input matrix and a second output matrix. The second input matrix includes stator core and winding losses, coolant outlet flow rate and temperature, and stator core and winding temperature. The second output matrix includes the temperature change rate of the stator core and winding and the temperature change rate of the coolant. The second data matrix is ​​used as training data to train the preset model.

[0013] Optionally, the method for constructing the motor temperature model further includes: The temperature prediction model is embedded into a nonlinear model predictive controller, with the temperature of the motor components as the state variable and the coolant flow rate as the control variable, to establish a motor flow control model with temperature constraints.

[0014] Secondly, the present invention provides a motor temperature model construction device, comprising: The first module is used to obtain simulation data of the motor based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model; The second module is used to train the preset model based on the simulation data and establish a temperature prediction model, which is used to predict the motor temperature.

[0015] Thirdly, the present invention provides an electronic device, including a memory and a processor; The memory is used to store computer programs; The processor is configured to implement the motor temperature model construction method as described in the first aspect when executing the computer program.

[0016] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the motor temperature model construction method as described in the first aspect.

[0017] The beneficial effects of the motor temperature model construction method of the present invention are as follows: the simulation data of the motor is obtained based on the one-dimensional thermal network model and the three-dimensional fluid thermal coupling model, and then the temperature prediction model is established based on the one-dimensional and three-dimensional simulation data. The prediction model is driven by data at different levels, which can avoid a large number of bench tests and lengthy three-dimensional simulation calculations, ensuring both fast response and high accuracy. This not only significantly shortens the design cycle and reduces costs, but also enables rapid and accurate prediction of the temperature of key motor components under complex working conditions. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the method for constructing a motor temperature model according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the process for obtaining simulation data according to an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the process of constructing a one-dimensional heat network model according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the process for constructing a three-dimensional fluid thermal coupling model according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the process for establishing a temperature prediction model according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the process for training the preset model according to an embodiment of the present invention. Figure 1 ; Figure 7 This is a schematic diagram of the process for training the preset model according to an embodiment of the present invention. Figure 2 ; Figure 8 This is a system architecture diagram of the motor temperature model construction device according to an embodiment of the present invention; Figure 9 This is a system architecture diagram of an electronic device according to an embodiment of the present invention; Figure 10 This is a schematic diagram of the thermal network model of a permanent magnet synchronous oil-cooled motor according to an embodiment of the present invention; Figure 11 This is a schematic diagram of the temperature results of different components in a one-dimensional simulation according to an embodiment of the present invention; Figure 12 This is a three-dimensional simulation temperature curve of an embodiment of the present invention; Figure 13 This is a schematic diagram of the neural network structure according to an embodiment of the present invention; Figure 14 This is a schematic diagram of regression error in an embodiment of the present invention; Figure 15 This is a schematic diagram comparing the results of one-dimensional simulation and prediction models in an embodiment of the present invention; Figure 16 This is a schematic diagram comparing the 3D simulation and model-predicted temperature of different components in an embodiment of the present invention; Figure 17 This is a schematic diagram of the motor flow control model according to an embodiment of the present invention. Detailed Implementation

[0019] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0020] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0021] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0022] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0023] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0024] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for constructing a motor temperature model, comprising: S100: Simulation data of the motor is obtained based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model.

[0025] Specifically, a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model were constructed respectively, and simulation data of the motor were obtained through simulation.

[0026] S200: The preset model is trained based on the simulation data to establish a temperature prediction model, which is used to predict the motor temperature.

[0027] Specifically, the preset model is trained based on simulation data to establish a one-dimensional simulation-driven temperature prediction model and a three-dimensional simulation-driven temperature prediction model for predicting motor temperature. The one-dimensional simulation-driven temperature prediction model can complete the simulation results obtained in a few seconds that would take several hours in a one-dimensional simulation, quickly verifying the temperature rise performance under complex working conditions. The three-dimensional simulation-driven temperature prediction model can obtain high-precision temperature results of motor components with time series in a short time, accelerating the acquisition of high-precision temperature. While ensuring accuracy, it greatly reduces simulation time. Moreover, when new simulation conditions appear in the project, it can avoid the large amount of time spent on three-dimensional CFD (Computational Fluid Dynamics) calculations and can directly output reliable temperature results. It can also be used for temperature control research on the whole machine.

[0028] In this embodiment, simulation data of the motor is obtained based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model. Then, a temperature prediction model is established based on the one-dimensional and three-dimensional simulation data. By using data at different levels to drive the prediction model, a large number of bench tests and lengthy three-dimensional simulation calculations can be avoided, ensuring both rapid response and high accuracy. This not only significantly shortens the design cycle and reduces costs, but also enables rapid and accurate prediction of the temperature of key motor components under complex operating conditions.

[0029] Optionally, the acquisition of motor simulation data based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model includes: S110: Construct the one-dimensional thermal network model, input component losses under different operating conditions into the one-dimensional thermal network model for simulation, and obtain the first temperature data of multiple locations in the motor.

[0030] Specifically, in combination Figure 2 , Figure 10 and Figure 11 As shown, based on the law of conservation of energy and the lumped parameter method, the structure of the permanent magnet synchronous oil-cooled motor is abstracted into an electrothermal resistance-thermal capacity network in a one-dimensional thermal simulation software. The motor component losses, speed and coolant flow rate under different operating conditions (such as WLTC, a mixed operating condition combining continuous and random operating conditions) are input. The operating time is 2573.2s, the time step is 0.1s, and the simulation takes 3 hours. The temperature and temperature change rate data of each component of the motor (such as stator core teeth and yoke, left-right end windings, overall winding, rotor core and magnet) are obtained.

[0031] Among them, component losses include stator copper losses, stator core tooth and yoke losses, magnet losses, rotor core losses, and wind-wear losses calculated from the rotational speed. The loss portion is defined by a table containing time series, and can input losses for various operating conditions such as WLTC, continuous acceleration and deceleration, and random conditions.

[0032] The motor cooling includes air cooling and oil cooling. Air cooling includes heat conduction through the air outside the motor housing and convection heat transfer around the rotor. Oil cooling includes heat dissipation through the stator core oil channels and spray cooling through the end windings. The oil cooling flow rate is defined by a table containing a time series, and multiple flow rates can be input to achieve flow control.

[0033] S120: Construct the three-dimensional fluid thermal coupling model, input component losses under different operating conditions into the three-dimensional fluid thermal coupling model for simulation, and obtain second temperature data at multiple locations in the motor, wherein the simulation data includes the first temperature data and the second temperature data.

[0034] Specifically, in combination Figure 2 and Figure 12 As shown, based on multiphysics coupling solution and conjugate heat transfer technology, a stator fluid thermal coupling model of a permanent magnet synchronous oil-cooled motor is built in a three-dimensional simulation software. The mixed working conditions are also input together, and the temperature and temperature change rate of the stator core and winding are collected. The winding part is divided into three parts: crown end, welded end and middle section.

[0035] In this optional embodiment, by combining one-dimensional thermal network simulation and three-dimensional fluid thermal coupling simulation to obtain temperature and temperature change rate data, it is possible to quickly cover multiple operating conditions and obtain high-precision results while avoiding a large number of experiments, achieving a complementarity between speed and accuracy, and providing reliable data support for subsequent prediction model training.

[0036] Optionally, constructing the one-dimensional heat network model includes: S111: Divide the motor into multiple hot nodes.

[0037] Specifically, in combination Figure 3 As shown, based on the lumped parameter method, the motor structure is abstracted into multiple thermal nodes (such as stator windings). Each node is characterized by thermal capacity for thermal inertia and by thermal resistance for inter-node thermal resistance.

[0038] S112: Set the thermal capacity and thermal resistance according to the component materials and structural parameters of the motor, and establish the conduction and convection relationship between the thermal nodes.

[0039] Specifically, in combination Figure 3 As shown, the heat capacity and thermal resistance are set according to the component material and structural parameters of the motor. For example, heat capacity parameters are set for each hot node (based on material parameters and structural parameters), and thermal resistance is established according to the geometric dimensions between nodes, the thermal conductivity of the material, and the contact interface conditions. Taking the conduction thermal resistance as an example, it can be directly calculated from the material and geometric dimensions. However, the convective thermal resistance, in addition to geometric parameters, must also consider the fluid and flow conditions. Therefore, the specific type and value of the thermal resistance can be determined by the conduction and convection relationship between hot nodes. At the same time, the oil passages and spray paths of the cooling medium can be modeled as a convective thermal resistance network coupled with the corresponding hot nodes.

[0040] S113: Input the various losses of the motor as heat sources into the corresponding heat nodes to construct the one-dimensional heat network model.

[0041] Specifically, in combination Figure 3As shown, various losses in motor operation are released as heat. These losses are mapped to different thermal nodes as input power. Therefore, various losses of the motor (such as copper loss, iron loss, magnet loss and mechanical loss) can be used as heat sources input to the corresponding thermal nodes. Based on the energy balance equation of each thermal node, a one-dimensional thermal network model composed of thermal resistance and thermal capacity is established (the heat transfer process of the motor is abstracted into a network composed of thermal resistance and thermal capacity, where loss determines heat input, thermal capacity determines the temperature rise rate, and thermal resistance determines heat diffusion). This model is used to output the temperature and temperature change rate of each node under different operating conditions.

[0042] In this optional embodiment, by dividing the motor into hot nodes and establishing an electrothermal resistance-thermal capacity network, the temperature rise characteristics of multiple components under different operating conditions can be quickly simulated, significantly reducing the calculation time and facilitating the rapid iteration of early structural design and cooling schemes.

[0043] Optionally, constructing the three-dimensional fluid thermal coupling model includes: S121: Establish a geometric model of the stator and windings of the motor, and discretize the geometric model to form a mesh model.

[0044] Specifically, in combination Figure 4 As shown, a geometric model of the motor stator and windings is established in a 3D simulation software. The geometric model determines the actual shape and boundary range of the motor stator and windings. Then, the geometric region is discretized and divided into a finite number of mesh elements.

[0045] S122: Establish the temperature field and flow field based on the mesh model to construct the three-dimensional fluid thermal coupling model.

[0046] Specifically, in combination Figure 4 As shown, on the established mesh, fluid flow (coolant velocity field, pressure field) and heat transfer (solid heat conduction, fluid convection, interface conjugate heat transfer) are coupled and solved to construct a three-dimensional fluid thermal coupling model. Through multi-physics coupling, the flow field and temperature field distribution inside the motor are finely characterized, which can accurately reflect the details of complex geometric structure, oil passage layout, contact thermal resistance and other details.

[0047] In this optional embodiment, by establishing a geometric and mesh model of the stator and windings and performing fluid-thermal coupling simulation, the local temperature distribution and hot spot location of the stator core and windings can be accurately captured, providing a high-precision basis for optimizing oil circuit design and cooling efficiency.

[0048] Optionally, training the preset model based on the simulation data to establish a temperature prediction model includes: S210: Train the preset model based on the first temperature data to establish a one-dimensional simulation-driven temperature prediction model.

[0049] Specifically, in combination Figure 5 As shown, the component temperature and temperature change rate data obtained by one-dimensional thermal network simulation are used as training samples to train the preset model and establish a one-dimensional simulation-driven temperature prediction model. The one-dimensional simulation-driven temperature prediction model is suitable for the pre-research stage of motor R&D. It can quickly determine whether the motor structure design and cooling strategy are reasonable and can cover a large number of working conditions for thermal performance verification in a short time.

[0050] S220: Train the preset model based on the second temperature data to establish a three-dimensional simulation-driven temperature prediction model.

[0051] Specifically, in combination Figure 5 As shown, the temperature and temperature change rate data of the stator core and windings obtained by three-dimensional fluid-thermal coupling simulation are used as training samples to train the preset model and establish a three-dimensional simulation-driven temperature prediction model. The three-dimensional simulation-driven temperature prediction model is suitable for the verification stage of motor R&D, and obtains high-precision temperature distribution results as the basis for design finalization. It is used to study local hot spot problems, such as the temperature rise risk in difficult areas such as the stator winding ends and oil flow dead zones. The one-dimensional prediction model can be used for rapid iteration first, and then the three-dimensional prediction model can be used for high-precision confirmation and control strategy calibration.

[0052] In this optional embodiment, by training temperature prediction models based on one-dimensional and three-dimensional simulation data respectively, it is possible to simultaneously obtain a fast-responding one-dimensional prediction model and a highly accurate three-dimensional prediction model, which can meet the needs of motor thermal management research at different stages and under different requirements.

[0053] Optionally, training the preset model based on the first temperature data includes: S211: Construct a first data matrix based on the first temperature data, wherein the first data matrix includes a first input matrix and a first output matrix, the first input matrix includes motor component losses, motor speed, coolant flow rate, coolant temperature and the temperature of each component of the motor, and the first output matrix includes the temperature change rate of each component of the motor and the temperature change rate of the coolant.

[0054] Specifically, in combination Figure 6 As shown, a first data matrix is ​​constructed based on the first temperature data. The first data matrix includes a first input matrix and a first output matrix. The first input matrix is ​​a matrix containing time-series features composed of 15 variables, including component wear, rotational speed, coolant flow rate and temperature, and the temperature of each component. The first output matrix is ​​a matrix containing time-series features composed of 8 variables, including the temperature change rate of each component and the coolant.

[0055] Among them, combined Figure 13As shown, a temperature prediction model can be built using a neural network. The input preprocessing layer uses symmetric normalization to process the parameters, achieving automatic scaling and enhancing data stability. The output layer performs inverse normalization. The intermediate layers consist of one hidden layer and one output layer. The hidden layer has 15 neurons with the hyperbolic tangent activation function, and the output layer has 8 neurons with the linear activation function. The first data matrix has 25,754 rows, divided into training, validation, and test sets with proportions of 70%, 15%, and 15%, respectively. The objective function is MSE, and the training algorithm chosen is the LM algorithm. After 1000 epochs of training, the regression error of the data is as follows: Figure 14 As shown in the figure, the training set, validation set, and test set all exhibit high regression accuracy. For the trained neural network model, a temperature integral model is constructed. Inputting operating condition data, the simulated and model-predicted temperatures of the stator core, stator windings, rotor core, and magnets are compared as follows: Figure 15 As shown, the temperature integral model can accurately fit the temperature time-series characteristics of the simulation model.

[0056] S212: Use the first data matrix as training data to train the preset model.

[0057] Specifically, in combination Figure 6 As shown, the first data matrix is ​​used as training data to train the neural network.

[0058] In this optional embodiment, by using component losses, rotational speed, coolant parameters, and component temperature as inputs and temperature change rate as output for training, the one-dimensional prediction model can comprehensively reflect the motor's operating status and accurately capture temperature rise dynamics, maintaining high prediction accuracy even under complex operating conditions.

[0059] Optionally, training the preset model based on the second temperature data includes: S221: Construct a second data matrix based on the second temperature data, wherein the second data matrix includes a second input matrix and a second output matrix, the second input matrix includes stator core and winding losses, coolant outlet flow rate and temperature, and stator core and winding temperature, and the second output matrix includes the temperature change rate of the stator core and winding and the temperature change rate of the coolant.

[0060] Specifically, in combination Figure 7As shown, a second data matrix was constructed based on the second temperature data. This second data matrix includes a second input matrix and a second output matrix. The second input matrix consists of 10 variables: stator core and winding losses, coolant outlet flow rate and temperature, and stator core and winding temperatures. The second output matrix represents the rate of change of stator core and winding temperatures and the rate of change of coolant temperatures. During training, the neural network has 10 neurons in its hidden layer, with the hyperbolic tangent activation function, and 5 neurons in its output layer, with a linear activation function. The collected data matrix has 13,000 rows, divided into training, validation, and test sets at 70%, 15%, and 15% respectively. The objective function is MSE, and the LM algorithm was selected for training. After 39 epochs of training, the datasets converged. The MSE values ​​for the three datasets were 0.00128, 0.00162, and 0.00188, with corresponding correlation coefficients R of 0.988, 0.986, and 0.982, indicating high overall accuracy. A temperature integral model was built using the trained neural network, and the results of different component models and simulations were compared as follows. Figure 16 As shown, the temperature change trend is consistent overall. The prediction model can capture the rapid temperature change trend of motor components under complex working conditions very well, with a maximum temperature error of <3℃, and the accuracy meets engineering requirements.

[0061] S222: Use the second data matrix as training data to train the preset model.

[0062] Specifically, in combination Figure 7 As shown, the second data matrix is ​​used as training data to train the neural network.

[0063] In this optional embodiment, by taking the loss and temperature parameters of the stator core and windings as inputs and the temperature change rate of the stator and coolant as outputs for training, the three-dimensional prediction model can focus on key components, accurately reflect the local temperature change trend, and achieve high-precision dynamic prediction.

[0064] Optionally, the method for constructing the motor temperature model further includes: The temperature prediction model is embedded into a nonlinear model predictive controller, with the temperature of the motor components as the state variable and the coolant flow rate as the control variable, to establish a motor flow control model with temperature constraints.

[0065] Specifically, in combination Figure 17As shown, the transparent module on the left is a nonlinear MPC controller. Based on the constructed neural network model, the state variables are the temperatures of different components, and the control variable is the coolant flow rate. The middle section represents the motor component losses, with the losses being: I (rotor iron loss), M (magnet loss), T (stator tooth loss), S (stator yoke loss), C (winding copper loss), and R (motor speed). The blue module on the right is a neural network temperature prediction model. The output is the component temperature change rate, and the output is used to obtain the temperature result through an integrator module.

[0066] In the MPC controller, different effects can be achieved by setting different constraints and objective functions: (1) MPC uses the motor thermal model to predict the temperature rise trend in the next few sampling periods in advance, and starts or enhances the cooling system in advance before the actual temperature reaches the peak, so as to prevent over-temperature protection from being triggered or performance from being degraded; (2) Under the premise of meeting the temperature rise constraint, MPC can finely adjust the cooling flow rate to avoid unnecessary cooling energy consumption; (3) The MPC framework inherently supports multi-objective multi-variable control, integrating all relevant actuators and controlled temperatures into one model.

[0067] The input of the nonlinear MPC controller consists of three parts: (1) eight state variables, namely the temperatures of different components and coolant, provided by the temperature calculated by the integrator; (2) eight reference values, namely the preset reference values, which correspond to the target temperature values ​​of the components in this embodiment, and can be used to control the specified components by setting weights in the controller; (3) seven control variables from the previous step, which are input by the unit time delay module from the component loss and the coolant flow rate in the control variables output by the MPC; (4) seven output variables are the control variables calculated by the MPC, including the calculated component loss and coolant flow rate. Since only the flow rate needs to be controlled in this embodiment, this purpose is achieved by setting the weight of the component loss part in the control variables to 0. The range of the control variables is set to the coolant flow rate range, i.e., [5, 15]. The state function of the MPC module is set to the function that calculates the state variables required by the corresponding controller using the trained neural network model, and the output function is set to the output of the state variables. The sampling time of the MPC module is consistent with the simulation at 0.1s, the prediction domain length is 20, the control domain length is 2, and the control variables can track the target changes well.

[0068] In this optional embodiment, by embedding the temperature prediction model into a nonlinear model predictive controller, with component temperature as the state variable and coolant flow rate as the control variable, the present invention can adjust the cooling in advance before the temperature reaches the peak, avoid overheating and derating operation, and optimize cooling energy consumption under the constraint conditions, thereby improving the safety and energy efficiency of thermal management.

[0069] like Figure 8 As shown, an embodiment of the present invention provides a motor temperature model construction device 800, comprising: The first module 810 is used to obtain simulation data of the motor based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model; The second module 820 is used to train the preset model based on the simulation data and establish a temperature prediction model, which is used to predict the motor temperature.

[0070] like Figure 9 As shown, an electronic device 900 provided in this embodiment of the invention includes a memory 920 and a processor 910; the memory 920 is used to store a computer program; the processor 910 is used to implement the motor temperature model construction method as described above when the computer program is executed.

[0071] Alternatively, an electronic device 900 includes a memory 920 and a processor 910 coupled to the memory 920; the memory 920 is configured to store a computer program; and the processor 910 is configured to perform the following operations when the computer program is executed: Simulation data of the motor were obtained based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model. The preset model is trained based on the simulation data to establish a temperature prediction model, which is used to predict the motor temperature.

[0072] This invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the motor temperature model construction method described above.

[0073] Alternatively, a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the following operations: Simulation data of the motor were obtained based on a one-dimensional thermal network model and a three-dimensional fluid thermal coupling model. The preset model is trained based on the simulation data to establish a temperature prediction model, which is used to predict the motor temperature.

[0074] The present invention will now be described an electronic device 900 that can serve as a server or client of the present invention, which is an example of a hardware device that can be applied to various aspects of the present invention. Electronic device 900 is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 900 can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0075] Electronic device 900 includes a computing unit that can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) or a computer program loaded from a storage unit into random access memory (RAM). The RAM may also store various programs and data required for device operation. The computing unit, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0076] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc. In this application, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention according to actual needs. Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units can be implemented in hardware or as software functional units.

[0077] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method of constructing a temperature model of an electric machine, characterized by, The method comprises the steps of: obtaining simulation data of the motor based on a one-dimensional thermal network model and a three-dimensional fluid-thermal coupling model; training a preset model according to the simulation data to establish a temperature prediction model, wherein the temperature prediction model is used to predict the temperature of the motor.

2. The motor temperature model building method of claim 1, wherein, The step of obtaining simulation data of the motor based on a one-dimensional thermal network model and a three-dimensional fluid-thermal coupling model comprises the steps of: constructing the one-dimensional thermal network model, inputting component losses under different working conditions into the one-dimensional thermal network model for simulation to obtain first temperature data of multiple positions in the motor; constructing the three-dimensional fluid-thermal coupling model, inputting component losses under different working conditions into the three-dimensional fluid-thermal coupling model for simulation to obtain second temperature data of multiple positions in the motor, wherein the simulation data comprises the first temperature data and the second temperature data.

3. The motor temperature model building method of claim 2, wherein, The step of constructing the one-dimensional thermal network model comprises the steps of: dividing the motor into multiple thermal nodes; setting thermal capacity and thermal resistance according to component materials and structural parameters of the motor, and establishing conduction and convection relationships between the thermal nodes; inputting various losses of the motor as heat sources into corresponding thermal nodes to construct the one-dimensional thermal network model.

4. The motor temperature model building method of claim 2, wherein The step of constructing the three-dimensional fluid-thermal coupling model comprises the steps of: establishing a geometric model of a stator and windings of the motor, discretizing the geometric model to form a grid model; establishing a temperature field and a flow field according to the grid model to construct the three-dimensional fluid-thermal coupling model.

5. The motor temperature model building method of claim 2, wherein The step of training a preset model according to the simulation data to establish a temperature prediction model comprises the steps of: training the preset model according to the first temperature data to establish a one-dimensional simulation-driven temperature prediction model; training the preset model according to the second temperature data to establish a three-dimensional simulation-driven temperature prediction model.

6. The motor temperature model building method of claim 5, wherein, The step of training the preset model according to the first temperature data comprises the steps of: constructing a first data matrix according to the first temperature data, wherein the first data matrix comprises a first input matrix and a first output matrix, the first input matrix comprises motor component losses, motor speed, cooling liquid flow, cooling liquid temperature, and temperatures of motor components, and the first output matrix comprises temperature change rates of motor components and a temperature change rate of the cooling liquid; training the preset model by taking the first data matrix as training data.

7. The motor temperature model building method of claim 5, wherein, The step of training the preset model according to the second temperature data comprises the steps of: constructing a second data matrix according to the second temperature data, wherein the second data matrix comprises a second input matrix and a second output matrix, the second input matrix comprises stator core and winding losses, cooling liquid outlet flow and temperature, and stator core and winding temperatures, and the second output matrix comprises temperature change rates of the stator core and the windings and a temperature change rate of the cooling liquid; training the preset model by taking the second data matrix as training data.

8. The method of claim 1 to 7, wherein The method further comprises the step of: embedding the temperature prediction model into a nonlinear model predictive controller, taking motor component temperatures as state variables and taking cooling liquid flow as a control variable to establish a motor flow control model with temperature constraints.

9. A motor temperature model construction apparatus characterized by comprising: The method comprises the steps of: The first module is configured to acquire simulation data of the motor based on a one-dimensional thermal network model and a three-dimensional fluid-thermal coupling model. The second module is configured to train a preset model according to the simulation data, and establish a temperature prediction model, which is configured to predict the motor temperature.

10. An electronic device, comprising: comprising a memory and a processor; The memory is configured to store a computer program. The processor is configured to implement the motor temperature model construction method according to any one of claims 1 to 8 when executing the computer program.

11. A computer readable storage medium, characterized in that, The storage medium has a computer program stored thereon, and the computer program, when executed by a processor, implements the motor temperature model construction method according to any one of claims 1 to 8.

Images