Electromechanical composite transmission IGBT junction temperature prediction method based on NODE reduced-order model

Abstract

The invention discloses an electromechanical composite transmission IGBT junction temperature prediction method based on a NODE reduced-order model. A thermal simulation model is constructed according to a real physical model mechanism, a material and a thermal transfer principle of an IGBT; in the constructed thermal simulation model, performing simulation calculation according to the monitored time sequence data to obtain the junction temperature of the IGBT corresponding to the time change, jointly establishing a data set for training and testing with the monitored data, establishing a NODE neural network order reduction model comprising an input layer, a hidden layer and an output layer, and training the NODE neural network order reduction model by using the data set; and inputting the IGBT related time sequence data into the constructed NODE neural network reduced-order model for prediction to obtain the time sequence data of the junction temperature. According to the method, the IGBT junction temperature can be predicted in real time, the calculation efficiency is high, the interpretability is high, the long-term dependence problem can be better solved, and the method is suitable for complex working conditions.

Description

Background Art

[0002] Insulated gate bipolar transistor (IGBT) modules are crucial components in the field of modern power electronics and are widely used in areas such as electric vehicles and industrial motor drives. IGBT modules combine the advantages of metal-oxide-semiconductor field-effect transistors (MOSFETs) and bipolar junction transistors (BJTs), featuring high efficiency, powerful performance, and fast switching capabilities. However, one of the most important factors affecting the performance, lifespan, and reliability of IGBT modules is the junction temperature. The junction temperature of an IGBT directly affects its operating efficiency, failure rate, and service life. Excessive junction temperature can not only lead to thermal runaway and degradation of semiconductor materials but may also cause device failures. Therefore, accurately and real-time monitoring the junction temperature of IGBT modules is crucial for ensuring their normal operation.

[0003] Currently, there are two methods for monitoring the IGBT junction temperature. One is the direct measurement method based on temperature sensors, and the other is the calculation method based on thermal models. The calculation method based on thermal models can relatively accurately monitor the change of junction temperature, but the computational amount is large, making it difficult to meet the requirements of real-time monitoring.

[0004] Traditional junction temperature monitoring methods mainly rely on the direct measurement of temperature sensors and the indirect estimation of thermal models. The direct measurement method usually uses temperature sensors such as thermocouples or resistance temperature detectors (RTDs), installed at different positions of the IGBT module to provide local temperature data. However, a single sensor often has difficulty comprehensively reflecting the temperature distribution of the entire module, and the sensor data may be affected by external environmental conditions (such as ambient temperature or air flow).

[0005] In contrast, thermal models provide a more comprehensive solution. These models can predict the change of junction temperature under different operating conditions by simulating the material properties, power losses, and heat transfer characteristics of IGBT modules. Although the thermal network model can accurately predict the junction temperature, its computational complexity is high and the calculation speed is slow, making it difficult to meet the requirements of real-time monitoring. Taking corresponding thermal management measures based on the junction temperature monitored by the thermal model has a certain time delay and cannot be adjusted dynamically in a timely manner. In addition, although the existing prediction methods based on BP neural networks and RBF neural networks can improve the calculation speed, they have limitations in dealing with time series data and dynamic changes and cannot well adapt to real-time changing working conditions.

[0006] Traditional deep neural networks for predicting junction temperature need to save the intermediate activation values in each layer, while neural differential equation networks only need to record the state of the system and the dynamics of how it changes over time. Neural differential equation networks avoid storing the intermediate results of each layer and are generally more efficient in memory usage than traditional neural network junction temperature prediction. Predicting junction temperature by this method requires less memory.

[0007] Ordinary differential equations provide a more intuitive perspective on dynamic systems, which can help us understand how the predicted junction temperature of the network evolves step by step over time. Compared with traditional neural networks, the operations of each layer are often black-box and difficult to interpret. Neural differential equation networks may provide better interpretability between the predicted junction temperature and the state characteristics of the system itself through continuous models.

[0008] When predicting the junction temperature, the neural differential equation neural network updates the hidden state by solving ordinary differential equations, rather than performing calculations at each time step like traditional neural networks. This enables the neural differential equation network to handle long-term dependence problems in predicting the junction temperature, effectively capture the continuity and dynamics in the time series, and more stably process long time series data of the junction temperature and its system state changing over time, avoiding the problems of gradient vanishing and gradient explosion when traditional neural networks process long sequences, and ensuring the accuracy of real-time prediction of the junction temperature. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to provide a method for predicting the IGBT junction temperature of an electromechanical composite drive based on a NODE reduced-order model in view of the deficiencies in the above-mentioned prior art, so as to solve the technical problems of high computational complexity, large memory requirements of the current method, and difficulty in applying in resource-constrained environments.

[0010] The present invention adopts the following technical solutions: A method for predicting the IGBT junction temperature of an electromechanical composite drive based on a NODE reduced-order model includes the following steps: Construct a thermal simulation model according to the true physical model mechanism, materials and heat transfer principle of the IGBT; In the constructed thermal simulation model, according to the monitored time series data, simulate and calculate the junction temperature of the IGBT corresponding to the time change, jointly establish a data set for training and testing with the monitored data, establish a NODE neural network reduced-order model including an input layer, a hidden layer and an output layer, and use the data set to train the NODE neural network reduced-order model; Input the time series data related to the IGBT into the constructed NODE neural network reduced-order model for prediction to obtain the time series data of the junction temperature.

[0011] Preferably, use the power loss of the IGBT as the heat source input of the thermal simulation model; construct the heat capacity part of the thermal model according to the size, thickness, materials and their physical properties of each layer structure from the chip to the radiator inside the IGBT; construct the heat dissipation part according to the convective heat dissipation from the IGBT housing to the environment and the flow rate and temperature parameters of the radiator water cooling; in the constructed thermal simulation model, simulate and calculate the junction temperature of the IGBT corresponding to the time change according to the relevant time series data.

[0012] Preferably, the NODE neural network reduced-order model includes: Input layer: responsible for receiving data samples and passing them to the hidden layer; Hidden layer: extracts effective features from the input data through learning and provides support for the calculation of the output layer; Output layer: through linear transformation, maps the internal representation of the model to a continuous value, uses an ODE solver to calculate the change of the hidden state over time according to the mapped value, and obtains the predicted junction temperature.

[0013] Preferably, the data samples include power consumption and heat dissipation information.

[0014] Preferably, the hidden layer includes a first sub-hidden layer and a second sub-hidden layer. Each sub-hidden layer performs a non-linear transformation on the data through weighted summation and an activation function; the first sub-hidden layer enables the neural layer to model the non-linear relationship between input features; the second sub-hidden layer is used to extract high-level features from the input data and provide support for the final prediction result.

[0015] Preferably, each neuron in the first sub-hidden layer receives a weighted input of power consumption and heat dissipation information, and sums the weighted inputs; the sum result is then non-linearly transformed through the tanh activation function to output the value of the neuron; Each neuron in the second sub-hidden layer receives the output of the first sub-hidden layer, performs weighted summation, and is non-linearly transformed again through the tanh activation function.

[0016] Preferably, the data set is divided into a training set and a test set according to a set ratio, the training set is larger than the test set. The samples in the training set are used to train the NODE neural network reduced-order model to obtain the predicted junction temperature. The predicted junction temperature is compared with the actual junction temperature to obtain a loss, and the parameters in the NODE neural network reduced-order model are corrected through backpropagation. The test set is used to test whether the trained NODE neural network reduced-order model meets the junction temperature prediction conditions.

[0017] Preferably, the relevant time series data includes motor speed, motor torque, cooling water flow rate and temperature for heat dissipation.

[0018] Preferably, the collected relevant time series data is smoothed to remove noise and outliers.

[0019] In a second aspect, an IGBT junction temperature prediction system based on a NODE neural network reduced-order model according to an embodiment of the present invention includes: A construction module that constructs a thermal simulation model according to the true physical model structure, materials and heat transfer principle of the IGBT; Training module: In the constructed thermal simulation model, based on the monitored time series data, the junction temperature of the IGBT corresponding to the time change is calculated through simulation, and a data set for training and testing is jointly established with the monitored data. A NODE neural network reduced-order model including an input layer, a hidden layer, and an output layer is established, and the NODE neural network reduced-order model is trained using the data set. Prediction module: Input the time series data related to the IGBT into the constructed NODE neural network reduced-order model for prediction to obtain the time series data of the junction temperature.

[0020] In a third aspect, a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned IGBT junction temperature prediction method for electromechanical composite transmission based on the NODE reduced-order model are implemented.

[0021] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium including a computer program. When the computer program is executed by a processor, the steps of the above-mentioned IGBT junction temperature prediction method for electromechanical composite transmission based on the NODE reduced-order model are implemented.

[0022] In a fifth aspect, a chip includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned IGBT junction temperature prediction method for electromechanical composite transmission based on the NODE reduced-order model are implemented.

[0023] In a sixth aspect, an embodiment of the present invention provides an electronic device including a computer program. When the computer program is executed by the electronic device, the steps of the above-mentioned IGBT junction temperature prediction method for electromechanical composite transmission based on the NODE reduced-order model are implemented.

[0024] Compared with the prior art, the present invention has at least the following beneficial effects: An IGBT junction temperature prediction method for electromechanical composite transmission based on the NODE reduced-order model. Aiming at the problems of large computational amount and certain delay in monitoring the junction temperature by the thermal simulation model, a method for predicting the IGBT junction temperature based on the neural ordinary differential equation (NODE) neural network reduced-order model is proposed, which can greatly improve the computational efficiency, requires less memory during the training and prediction of the junction temperature, has better interpretability, and can handle long-term dependence problems. The NODE network model describes the dynamic behavior of the neural network through differential equations, can process continuous time data, and is particularly suitable for the prediction of time series data. Compared with the traditional neural network model, the NODE neural network reduced-order model has significant advantages in dealing with long-term dependence problems and can better capture the dynamic characteristics of the IGBT junction temperature changing with time.

[0025] Furthermore, the model order reduction technology shortens the calculation time from minutes to milliseconds, meeting the real-time requirements of embedded systems; optimizing the radiator parameters can reduce the system energy consumption.

[0026] Furthermore, the NODE neural network reduced-order model has a closed-loop design with multi-source data fusion in the input layer, continuous-time feature evolution in the hidden layer, and dynamic feedback correction in the output layer. In the junction temperature prediction task, the calculation speed is improved. The combination of hidden layer feature decoupling and ODE dynamics equations makes the model decision-making process transparent. Low memory and low computing power requirements support edge deployment, providing a real-time and reliable thermal management solution for power electronic systems.

[0027] Furthermore, the calculation speed is improved to meet the real-time control requirements at the 50ms level. It provides a solution with both physical interpretability and data-driven adaptability for high-reliability power electronic systems.

[0028] Furthermore, the hidden layer structure with double tanh activation approximates complex thermal dynamics equations through phased nonlinear modeling and gradient stability optimization. In the IGBT junction temperature prediction, the double nonlinear transformation meets the real-time control requirements at the 50ms level for a single prediction. It provides an efficient and robust solution for the real-time thermal management of power electronic systems.

[0029] It can be understood that the beneficial effects of the second to sixth aspects above can be referred to the relevant descriptions in the first aspect above, and will not be elaborated here.

[0030] In summary, the present invention can predict the IGBT junction temperature in real time, has high calculation efficiency, strong interpretability, can better handle long-term dependence problems, and is applicable to complex working conditions.

[0031] Next, through the drawings and embodiments, the technical solutions of the present invention will be further described in detail. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments of the present application will be briefly introduced below. Obviously, the following described drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0033] Figure 1 is the overall flowchart of the present invention; Figure 2 is the structural diagram of the NODE neural network reduced-order model; Figure 3 is the schematic diagram of the computer device provided by an embodiment of the present invention; Figure 4Block diagram of an electronic device provided according to an embodiment of the present invention.

[0034] Among them, 60. Computer device; 61. Processor; 62. Memory; 63. Computer program; 600. Electronic device; 610. Processing unit; 620. Storage unit; 6201. Random access storage unit; 6202. Cache storage unit; 6203. Read-only storage unit; 6204. Program / utilities; 6205. Program module; 630. Bus; 640. Display unit; 650. Input / output interface; 660. Network adapter; 700. External device. Detailed implementation

[0035] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0036] In the description of the present invention, it should be understood that the terms "include" and "comprise" indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0037] It should also be understood that the terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0038] It should be further understood that the term " / and" used in the specification of the present invention and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in the present invention generally represents an "or" relationship between the preceding and following associated objects.

[0039] It should be understood that although the terms first, second, third, etc. may be used in the embodiments of the present invention to describe preset ranges, etc., these preset ranges should not be limited to these terms. These terms are only used to distinguish the preset ranges from each other. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0040] Depending on the context, as used herein, the word "if" can be interpreted as "when" or "while" or "in response to determining" or "in response to detecting". Similarly, depending on the context, the phrase "if determined" or "if detecting (stated condition or event)" can be interpreted as "when determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)".

[0041] Various structural schematic diagrams according to the disclosed embodiments of the present invention are shown in the drawings. These figures are not drawn to scale, where for the purpose of clear expression, some details are enlarged and some details may be omitted. The shapes of various regions and layers shown in the figures, as well as their relative sizes and positional relationships, are merely exemplary, and in practice, there may be deviations due to manufacturing tolerances or technical limitations. Those skilled in the art can additionally design regions / layers with different shapes, sizes, and relative positions according to actual requirements.

[0042] The present invention provides a method for predicting the IGBT junction temperature of an electromechanical composite drive based on a NODE reduced-order model. After using a Neural Ordinary Differential Equation (Neural ODE) network to learn and train the data of the thermal model, the reduction of the original thermal model is achieved, the speed of real-time junction temperature prediction is improved, the memory required for the trained reduced-order model to predict the junction temperature is smaller, it has better interpretability, can handle the long-term dependence problem of the model, ensure the prediction accuracy, and provide effective data support for further dynamic thermal management systems.

[0043] The technical problems that can be solved are as follows: 1. Traditional deep neural networks for predicting junction temperature need to save the intermediate activation values in each layer, while the Neural Ordinary Differential Equation network only needs to record the state of the system and the dynamics of how it changes over time. The Neural Ordinary Differential Equation network avoids storing the intermediate results of each layer and is generally more efficient in memory usage than traditional neural network junction temperature prediction. The memory required for predicting the junction temperature by this method is smaller.

[0044] 2. Ordinary differential equations provide a more intuitive perspective of the dynamic system, which can help us understand how the network evolves step by step over time. Compared with traditional neural networks, the operations of each layer are often black boxes and difficult to interpret. The Neural Ordinary Differential Equation network can provide better interpretability between the predicted junction temperature and the state characteristics of the system itself through a continuous model.

[0045] 3. Neural differential equation networks update hidden states by solving ordinary differential equations instead of computing at each time step as in traditional neural networks. This enables neural differential equation networks to handle long-term dependence problems in predicting junction temperature, effectively capture the continuity and dynamics in time series, and more stably process long time series data of junction temperature and its system state changing over time, avoiding the problems of vanishing gradients and exploding gradients when traditional neural networks process long sequences, and ensuring the accuracy of real-time prediction of junction temperature.

[0046] 4. By using a neural ordinary differential equation (Neural ODE) neural network to fit the dynamic behavior of the system, the computational load can be effectively reduced, and real-time monitoring and prediction of the junction temperature can be carried out based on some directly measurable data, providing effective data support for further dynamic thermal management systems.

[0047] Embodiment 1 Please refer to Figure 1 , a method for predicting the IGBT junction temperature based on the NODE reduced-order model of the present invention, comprising the following steps: S1. Record the relevant time series data detected during the operation of the IGBT, The relevant time series data includes motor speed, motor torque, radiator water-cooling flow rate and temperature, etc.; these data are collected in real time by sensors installed on the IGBT module and stored and processed according to the time series.

[0048] Data acquisition and processing: Collect the time series data that can be directly monitored during the operation of the IGBT, and obtain data such as motor speed, motor torque, radiator water-cooling flow rate and temperature in real time through sensors. Smooth the collected time series data to remove noise and outliers.

[0049] The introduction of time series data enables the NODE neural network reduced-order model to better capture the dynamic changes of the IGBT junction temperature, thereby realizing real-time prediction.

[0050] S2. Construct a thermal simulation model according to the true physical model structure, materials and heat transfer principle of the IGBT; The thermal simulation model is constructed by using Simcenter Amesim software according to the true physical model structure, materials and heat transfer principle of the IGBT. The power loss of the IGBT is used as the heat source input of the thermal simulation model; according to the size, thickness, materials and their physical properties of each layer structure from the chip to the radiator inside the IGBT, the heat capacity part of the thermal model is constructed; according to the convective heat dissipation from the IGBT housing to the environment, parameters such as the flow rate and temperature difference of the radiator water-cooling, the heat dissipation part is constructed.

[0051] S3. In the thermal simulation model constructed in step S2, the junction temperature of the IGBT corresponding to the time variation is calculated by simulation according to the monitored time series data, and a data set for training and testing is jointly established with the monitoring data. A NODE neural network reduced-order model including an input layer, a hidden layer, and an output layer is established and trained using the data set. Please refer to Figure 2 , the NODE neural network reduced-order model includes an input layer, a hidden layer, and an output layer, specifically as follows: Input layer: Responsible for receiving data samples, including information such as ambient temperature and power consumption. The role of the input layer is to pass these raw data to the hidden layer for further processing.

[0052] Hidden layer: It includes two sub-hidden layers. Each sub-hidden layer performs a non-linear transformation on the data through weighted summation and activation functions. The role of the hidden layer is to extract effective features from the input data through learning and provide support for the calculation of the output layer. Each neuron in the first sub-hidden layer receives the weighted inputs of the ambient temperature and power consumption, and sums the weighted inputs. The sum result is then non-linearly transformed through the tanh activation function (also known as tansig) to output the value of the neuron. This process enables this neural layer to model the non-linear relationship between the input features; each neuron in the second sub-hidden layer receives the output of the first sub-hidden layer, sums through weighted summation, and is non-linearly transformed again through the tanh activation function. The role of this layer is to further extract the high-level features in the input data and provide support for the final prediction result.

[0053] Output layer: The output layer is connected to all neurons of the second sub-hidden layer. Through linear transformation, the output layer maps the internal representation of the model to a continuous value. Using an ODE solver, the change in the hidden state over time is calculated based on the mapped value, and the predicted junction temperature is obtained.

[0054] The NODE neural network reduced-order model is trained based on the monitored data and the junction temperature data predicted by the thermal simulation model.

[0055] The actually monitored time series data and the junction temperature data of the IGBT calculated by the thermal simulation model are jointly used to form a data set for training and testing the NODE neural network reduced-order model. Each data sample corresponds to different power losses and heat dissipation conditions, and also includes the junction temperature simulated under this condition, and this junction temperature is used as the actual junction temperature.

[0056] The data set is divided into a training set and a test set according to a certain ratio. The training set should be larger than the test set. The training set is used to train the NODE neural network reduced-order model, and the test set is used to test whether the trained NODE neural network reduced-order model meets the junction temperature prediction conditions.

[0057] Use the samples in the training set to train the NODE neural network reduced-order model to obtain the predicted junction temperature. Compare the junction temperature predicted by the model with the actual junction temperature to obtain the loss, and correct the parameters in the neural network through backpropagation.

[0058] Optimize the parameters of the NODE reduced-order model through the backpropagation algorithm to gradually reduce the loss function; input the test set into the trained NODE neural network reduced-order model for prediction. When the accuracy of the predicted junction temperature and the actual junction temperature of the test set meets the requirements, complete the training and testing of the NODE neural network reduced-order model.

[0059] If the accuracy of the predicted junction temperature and the actual junction temperature of the test set does not meet the requirements, continue training until the accuracy requirement of the trained NODE neural network reduced-order model for testing is met, and complete the training and testing of the NODE neural network reduced-order model.

[0060] S4. Input the IGBT-related time series data obtained in step S1 into the NODE neural network reduced-order model constructed in step S3 for prediction to obtain the time series data of the junction temperature.

[0061] During the process of predicting the junction temperature of the IGBT by the NODE neural network reduced-order model, it can simulate the temperature dynamic characteristics of the IGBT. Taking time series data (such as motor speed, motor torque, radiator water-cooling flow rate and temperature, etc.) as input, the NODE neural network reduced-order model learns the differential equation of the junction temperature change, can realize the prediction of the final junction temperature, can also predict the entire temperature change trajectory, can fit the actual physical law of the IGBT thermal characteristics, and the prediction effect is more accurate.

[0062] Those skilled in the art can understand that various aspects of the present invention can be implemented as a system, method or program product. Therefore, various aspects of the present invention can be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or an implementation combining hardware and software aspects, which can be collectively referred to as "circuit", "module" or "platform" here.

[0063] Embodiment 2 The present invention provides an IGBT junction temperature prediction system for an electromechanical composite transmission based on a NODE neural network reduced-order model. This system can be used to implement the above-mentioned IGBT junction temperature prediction method for an electromechanical composite transmission based on a NODE reduced-order model. Specifically, the IGBT junction temperature prediction system for an electromechanical composite transmission based on a NODE neural network reduced-order model includes a construction module, a training module, and a prediction module.

[0064] Among them, the construction module constructs a thermal simulation model according to the real physical model structure, materials and heat transfer principle of the IGBT; Training module: In the constructed thermal simulation model, based on the monitored time series data, the junction temperature of the IGBT corresponding to the time change is simulated and calculated, and a data set for training and testing is jointly established with the monitoring data. A NODE neural network reduced-order model including an input layer, a hidden layer, and an output layer is established, and the NODE neural network reduced-order model is trained using the data set. Input layer: Responsible for receiving data samples, including information such as ambient temperature and power consumption. The role of the input layer is to transfer these raw data to the hidden layer for further processing.

[0065] Hidden layer: It includes two sub-hidden layers. Each sub-hidden layer performs a non-linear transformation on the data through weighted summation and activation functions. The role of the hidden layer is to extract effective features from the input data through learning and provide support for the calculation of the output layer. Each neuron in the first sub-hidden layer receives the weighted inputs of the ambient temperature and power consumption, and sums the weighted inputs. The summation result is then non-linearly transformed through the tanh activation function (also known as tansig) to output the value of the neuron. This process enables this neural layer to model the non-linear relationship between the input features; each neuron in the second sub-hidden layer receives the output of the first sub-hidden layer, performs weighted summation, and is again non-linearly transformed through the tanh activation function. The role of this layer is to further extract the high-level features in the input data and provide support for the final prediction result.

[0066] Output layer: The output layer is connected to all neurons of the second sub-hidden layer. Through linear transformation, the output layer maps the internal representation of the model to a continuous value. Using an ODE solver, the change of the hidden state over time is calculated based on the mapped value, and the predicted junction temperature is obtained.

[0067] The NODE neural network reduced-order model is trained based on the monitored data and the junction temperature data predicted by the thermal simulation model.

[0068] The actually monitored time series data and the junction temperature data of the IGBT calculated by the thermal simulation model are jointly used to form a data set for training and testing the NODE neural network reduced-order model. Each data sample corresponds to different power losses and heat dissipation conditions, and also includes the junction temperature simulated under this condition, and this junction temperature is used as the actual junction temperature.

[0069] The data set is divided into a training set and a test set according to a certain ratio. The training set should be larger than the test set. The training set is used to train the NODE neural network reduced-order model, and the test set is used to test whether the trained NODE neural network reduced-order model meets the junction temperature prediction conditions.

[0070] The samples in the training set are used to train the NODE neural network reduced-order model to obtain the predicted junction temperature. The predicted junction temperature is compared with the actual junction temperature to obtain the loss, and the parameters in the neural network are corrected through backpropagation.

[0071] The parameters of the NODE reduced-order model are optimized through the backpropagation algorithm to gradually reduce the loss function. When the accuracy of the predicted junction temperature and the actual junction temperature of the test set reaches the requirement, the training and testing of the NODE neural network reduced-order model are completed.

[0072] If the accuracy of the predicted junction temperature and the actual junction temperature of the test set does not meet the requirement, continue training until the accuracy requirement of the completed test of the NODE neural network reduced-order model is met, and the training and testing of the NODE neural network reduced-order model are completed.

[0073] The prediction module inputs the IGBT-related time series data into the constructed NODE neural network reduced-order model for prediction to obtain the time series data of the junction temperature.

[0074] Embodiment 3 The present invention provides a terminal device, which includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Graphics Processing Unit (GPU), Tensor Processing Unit (TPU), Digital Signal Processor (DSP), Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the electromechanical composite drive IGBT junction temperature prediction method based on the NODE reduced-order model, including: Construct a thermal simulation model based on the true physical model structure, materials, and heat transfer principle of the IGBT; in the constructed thermal simulation model, calculate the junction temperature of the IGBT corresponding to the time change according to the monitored time series data through simulation, establish a data set for training and testing together with the monitored data, establish a NODE neural network reduced-order model including an input layer, a hidden layer, and an output layer, and use the data set to train the NODE neural network reduced-order model; input the IGBT-related time series data into the constructed NODE neural network reduced-order model for prediction to obtain the time series data of the junction temperature.

[0075] Please refer to Figure 3 , the terminal device is a computer device. The computer device 60 of this embodiment includes: a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When the computer program 63 is executed by the processor 61, it implements the method for predicting the IGBT junction temperature of the electromechanical composite drive based on the NODE reduced-order model in the embodiment. To avoid repetition, it will not be elaborated here one by one. Alternatively, when the computer program 63 is executed by the processor 61, it implements the functions of each model / unit in the system for predicting the IGBT junction temperature of the electromechanical composite drive based on the NODE neural network reduced-order model in the embodiment. To avoid repetition, it will not be elaborated here one by one.

[0076] The computer device 60 can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art can understand that Figure 3 This is only an example of the computer device 60 and does not constitute a limitation on the computer device 60. It may include more or fewer components than shown in the figure, or combine some components, or different components. For example, the computer device may also include input / output devices, network access devices, buses, etc.

[0077] The so-called processor 61 may be a Central Processing Unit (CPU), or other general-purpose processors, Graphics Processing Unit (GPU), Tensor Processing Unit (TPU), Digital Signal Processor (DSP), Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0078] The memory 62 may be an internal storage unit of the computer device 60, such as the hard disk or memory of the computer device 60. The memory 62 may also be an external storage device of the computer device 60, such as a plug-in hard disk equipped on the computer device 60, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc.

[0079] Furthermore, the memory 62 may also include both the internal storage unit of the computer device 60 and the external storage device. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 may also be used to temporarily store data that has been output or will be output.

[0080] Please refer to Figure 4 , the terminal device is the electronic device 600, and the electronic device 600 is presented in the form of a general computing device. The components of the electronic device may include but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including the storage unit 620 and the processing unit 610), a display unit 640, etc.

[0081] Among them, the storage unit stores program code, and the program code can be executed by the processing unit 610, so that the processing unit 610 executes the steps according to various exemplary embodiments of the present invention described in the above method part of this specification. For example, the processing unit 610 may execute the steps as shown in Figure 1 .

[0082] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 6201 and / or a cache storage unit 6202, and may further include a read-only storage unit (ROM) 6203.

[0083] The storage unit 620 may also include a program / utilities 6204 having a set (at least one) of program modules 6205. Such program modules 6205 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment.

[0084] The bus 630 may represent one or more of several types of bus structures, including a storage unit bus or storage unit controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of a variety of bus structures.

[0085] The electronic device 600 may also communicate with one or more external devices 700 (such as a keyboard, a pointing device, a Bluetooth device, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or may communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (such as a router, a modem). Such communication may be carried out through the input / output interface 650. Also, the electronic device 600 may communicate with one or more networks (such as a local area network, a wide area network, and / or a public network, such as the Internet) through the network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 through the bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms, etc.

[0086] Example 4 The present invention also provides a storage medium, specifically a computer-readable storage medium, which is a memory device in a terminal device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and, of course, the extended storage medium supported by the terminal device. It can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device. The computer-readable storage medium provides a storage space, and the operating system of the terminal is stored in this storage space. Moreover, one or more instructions suitable for being loaded and executed by the processor are also stored in this storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that more specific examples of the computer-readable storage medium here include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0087] The computer-readable storage medium also includes data signals propagated in a baseband or as part of a carrier wave, in which the readable program code is carried. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable storage medium can also be any readable medium other than the readable storage medium, and this readable medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium can be transmitted using any suitable medium, including but not limited to wireless, wired, optical cable, radio frequency, etc., or any suitable combination of the above.

[0088] The program code for performing the operations of the present invention can be written in any combination of one or more programming languages. The programming languages include object-oriented programming languages - such as Java, C++, Python, etc., and also include conventional procedural programming languages - such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, executed as an independent software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user's computing device through any type of network, including a local area network or a wide area network, or can be connected to an external computing device (for example, by using an Internet service provider to connect through the Internet).

[0089] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the method for predicting the IGBT junction temperature based on the NODE reduced-order model in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor as follows: Construct a thermal simulation model according to the true physical model structure, materials, and heat transfer principles of the IGBT; in the constructed thermal simulation model, calculate the junction temperature of the IGBT corresponding to the time change by simulating the monitored time series data, jointly establish a data set for training and testing with the monitored data, establish a NODE neural network reduced-order model including an input layer, a hidden layer, and an output layer, and use the data set to train the NODE neural network reduced-order model; input the IGBT-related time series data into the constructed NODE neural network reduced-order model for prediction to obtain the time series data of the junction temperature.

[0090] The databases involved in the embodiments provided in the present application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., and is not limited thereto. The processors involved in the embodiments provided in the present application may be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., and is not limited thereto.

[0091] In summary, the present invention provides a method for predicting the IGBT junction temperature of an electromechanical composite drive based on a NODE reduced-order model, introducing a NODE neural network reduced-order model: using a neural network reduced-order model based on neural ordinary differential equations (Neural ODE) can process continuous-time data and is particularly suitable for predicting time series data; the NODE network reduced-order model describes the dynamic behavior of the neural network through differential equations and can better capture the dynamic characteristics of the IGBT junction temperature changing with time; it can predict the IGBT junction temperature in real time and is suitable for dynamically changing working conditions; the NODE network model significantly reduces the computational complexity and memory requirements and improves the prediction speed; the model based on differential equations has better interpretability and can better understand the prediction process of the model; the NODE neural network reduced-order model can better handle the long-term dependence problem in time series data and improve the prediction accuracy. It can process a variety of input parameters and is suitable for complex actual working conditions.

[0092] Those skilled in the art can clearly understand that, for the convenience and conciseness of description, only the above-mentioned division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiments can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working processes of the units and modules in the above system can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated here.

[0093] In the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0094] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in the present invention can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0095] In the embodiments provided by the present invention, it should be understood that the disclosed device / terminal and method can be implemented in other ways. For example, the device / terminal embodiments described above are only illustrative. For example, the division of the module or unit is only a logical function division. In actual implementation, there can be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in an electrical, mechanical or other form.

[0096] The unit described as a separated component may or may not be physically separated, and the component displayed as a unit may or may not be a physical unit, that is, it can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0097] In addition, in each embodiment of the present invention, each functional unit can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0098] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above-mentioned embodiment methods of the present invention, it can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0099] This application is described with reference to the flowcharts and / or block diagrams of methods, devices, and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0100] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device implements the functions in the process Figure 1One or more processes and / or boxes Figure 1 The functions specified in one or more boxes.

[0101] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 One or more processes and / or boxes Figure 1 The steps of the functions specified in one or more boxes.

[0102] The above content is only to illustrate the technical idea of the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention fall within the protection scope of the claims of the present invention.

Claims

1. A method for predicting the IGBT junction temperature of an electromechanical composite drive based on a NODE reduced-order model, characterized in that It includes the following steps: Construct a thermal simulation model according to the real physical model structure, materials, and heat transfer principle of the IGBT; In the constructed thermal simulation model, based on the monitored time series data, simulate and calculate the junction temperature of the IGBT corresponding to the time change, jointly establish a data set for training and testing with the monitored data, establish a NODE neural network reduced-order model including an input layer, a hidden layer, and an output layer, and use the data set to train the NODE neural network reduced-order model; Input the time series data related to the IGBT into the constructed NODE neural network reduced-order model for prediction to obtain the time series data of the junction temperature.

2. The IGBT junction temperature prediction method based on the NODE reduced-order model according to claim 1, characterized in that Use the power loss of the IGBT as the heat source input of the thermal simulation model; according to the dimensions, thicknesses, materials, and their physical properties of each layer structure from the chip to the radiator inside the IGBT, construct the heat capacity part of the thermal model; Construct the heat dissipation part according to the convective heat dissipation from the IGBT housing to the environment and the flow rate and temperature parameters of the radiator water cooling; In the constructed thermal simulation model, based on the relevant time series data, simulate and calculate the junction temperature of the IGBT corresponding to the time change.

3. The IGBT junction temperature prediction method based on the NODE reduced-order model according to claim 1, wherein, The NODE neural network reduced-order model includes: Input layer: Responsible for receiving data samples and passing them to the hidden layer; Hidden layer: Extract effective features from the input data through learning and provide support for the calculation of the output layer; Output layer: Through linear transformation, map the internal representation of the model to a continuous value, use an ODE solver to calculate the change of the hidden state over time according to the mapped value, and obtain the predicted junction temperature.

4. The IGBT junction temperature prediction method for the electromechanical composite drive based on the NODE reduced-order model according to claim 3, characterized in that The data samples include power consumption and heat dissipation information.

5. The IGBT junction temperature prediction method based on the NODE reduced-order model according to claim 3, characterized in that, The hidden layer includes a first sub-hidden layer and a second sub-hidden layer. Each sub-hidden layer performs a non-linear transformation on the data through weighted summation and an activation function; the first sub-hidden layer enables this neural layer to model the non-linear relationship between input features; The second sub-hidden layer is used to extract high-level features from the input data and provide support for the final prediction result.

6. The method for predicting the IGBT junction temperature of the electromechanical composite drive based on the NODE reduced-order model according to claim 5, wherein, Each neuron in the first sub-hidden layer receives the weighted input of power consumption and heat dissipation information, sums the weighted input, and then performs a non-linear transformation on the summation result through the tanh activation function to output the value of the neuron; Each neuron in the second sub-hidden layer receives the output of the first sub-hidden layer, performs weighted summation, and again performs a non-linear transformation through the tanh activation function.

7. The IGBT junction temperature prediction method for the electromechanical composite drive based on the NODE reduced-order model according to claim 1, wherein Divide the data set into a training set and a test set according to a set ratio, with the training set larger than the test set. Use the samples in the training set to train the NODE neural network reduced-order model to obtain the predicted junction temperature, compare the predicted junction temperature with the actual junction temperature to obtain the loss, and correct the parameters in the NODE neural network reduced-order model through backpropagation. The test set is used to test whether the trained NODE neural network reduced-order model meets the junction temperature prediction conditions.

8. The IGBT junction temperature prediction method based on the NODE reduced-order model according to claim 1, characterized in that, The relevant time series data includes motor speed, motor torque, radiator water cooling flow rate, and temperature.

9. The IGBT junction temperature prediction method for electromechanical composite drive based on the NODE reduced-order model according to claim 8, wherein, Perform smoothing processing on the collected relevant time series data to remove noise and outliers.

10. An IGBT junction temperature prediction system for electromechanical composite drive based on the reduced-order model of the NODE neural network, characterized in that, It includes: Construction module, which constructs a thermal simulation model according to the real physical model structure, materials, and heat transfer principle of the IGBT; Training module: In the constructed thermal simulation model, based on the monitored time series data, the junction temperature of the IGBT corresponding to the time change is simulated and calculated, and a data set for training and testing is jointly established with the monitoring data. A NODE neural network reduced-order model including an input layer, a hidden layer, and an output layer is established, and the NODE neural network reduced-order model is trained using the data set; Prediction module: The time series data related to the IGBT is input into the constructed NODE neural network reduced-order model for prediction to obtain the time series data of the junction temperature.

Images