Junction temperature prediction method and device for power device of motor controller of unmanned aerial vehicle, and electronic equipment

Abstract

The invention provides a junction temperature prediction method and device for a power device of a motor controller of an unmanned aerial vehicle and electronic equipment, and relates to the technical field of power electronic device detection, and the method comprises the steps: obtaining electrical characteristic parameters in the operation process of the power device of the motor controller of the unmanned aerial vehicle, operation condition information of a motor system, and corresponding environment parameters; inputting the electrical characteristic parameters and the environmental parameters into a residual life prediction model to obtain the residual life of the power device; and inputting the residual life, the operation condition information and the environmental parameters into the junction temperature prediction model to obtain the junction temperature of the power device. According to the invention, the residual life of the power device is predicted based on the electrical characteristic parameters and the environmental parameters through the residual life prediction model, and then the junction temperature of the power device is predicted based on the residual life, the operation condition information and the environmental parameters through the junction temperature prediction model. The accuracy and reliability of junction temperature prediction of the power device can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power electronic device detection, and in particular to a method, device and electronic equipment for predicting the junction temperature of a power device of a motor controller of an unmanned aerial vehicle. Background Art

[0002] UAVs are an important carrier of the low-altitude economy, and motor controllers are the core components of their power systems. As basic components for realizing power electronic conversion, various power devices have a key impact on the reliable and efficient operation of motor controllers. Excessively high junction temperature of power devices will increase the power loss of motor controllers, reduce system operating efficiency and control performance, and even endanger the reliable operation of the system in severe cases. Therefore, in order to ensure the efficient and reliable operation of UAV systems, it is necessary to accurately estimate the junction temperature of power devices during operation.

[0003] In the related art, the collected electrical parameters of the power device are usually input into a junction temperature prediction model, and the junction temperature of the power device is predicted by the junction temperature prediction model.

[0004] However, in the above-mentioned related technologies, the junction temperature of the power device is predicted only based on the electrical parameters of the power device, which reduces the accuracy and reliability of the prediction of the junction temperature of the power device. Summary of the invention

[0005] The present invention provides a method, device and electronic equipment for predicting the junction temperature of a power device of a motor controller of a drone, so as to solve the defect of reducing the accuracy and reliability of the prediction of the junction temperature of the power device in the prior art.

[0006] The present invention provides a method for predicting the junction temperature of a power device of a motor controller of a drone, comprising: Obtaining electrical characteristic parameters of the power device of the UAV motor controller during operation, operating condition information of the motor system of the UAV, and environmental parameters corresponding to the UAV; Inputting the electrical characteristic parameters and the environmental parameters into a remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model, wherein the remaining life prediction model is trained based on sample electrical characteristic parameters of sample power devices of sample UAV motor controllers and sample environmental parameters corresponding to the sample UAVs; The remaining life, the operating condition information and the environmental parameters are input into a junction temperature prediction model to obtain the junction temperature of the power device output by the junction temperature prediction model. The junction temperature prediction model is trained based on the sample remaining life of the sample power device, the sample operating condition information of the motor system of the sample UAV and the sample environmental parameters.

[0007] According to a method for predicting junction temperature of a power device of a motor controller of a drone provided by the present invention, the remaining life prediction model is trained based on the following method: Obtaining sample electrical characteristic parameters of sample power devices of the sample UAV motor controller and sample environmental parameters corresponding to the sample UAV; inputting the sample electrical characteristic parameters and the sample environmental parameters into an initial remaining life prediction model to obtain a predicted remaining life output by the initial remaining life prediction model; Based on the predicted remaining life and the remaining life labels of the sample power devices, model parameters of the initial remaining life prediction model are iteratively optimized to obtain the remaining life prediction model.

[0008] According to a method for predicting junction temperature of a power device of a motor controller of a drone provided by the present invention, the sample electrical characteristic parameters and the sample environmental parameters are input into an initial remaining life prediction model to obtain a predicted remaining life output by the initial remaining life prediction model, including: For each test, the sample electrical characteristic parameters and sample environmental parameters obtained in this test are input into the initial remaining life prediction model to obtain the predicted remaining life of the sample power device corresponding to this test output by the initial remaining life prediction model; The iterative optimization of the model parameters of the initial remaining life prediction model based on the predicted remaining life and the remaining life label of the sample power device to obtain the remaining life prediction model includes: Based on the predicted remaining life of the sample power devices corresponding to each test and the remaining life labels of the sample power devices corresponding to each test, the model parameters of the initial remaining life prediction model are iteratively optimized to obtain the remaining life prediction model.

[0009] According to a method for predicting junction temperature of a power device of a motor controller of a drone provided by the present invention, the method further comprises: Obtaining a total number of tests corresponding to an upper limit of the service life of the sample power device; For each test, based on the test sequence number representing the test and the total number of tests, the remaining life label corresponding to the test is determined.

[0010] According to a method for predicting the junction temperature of a power device of a motor controller of a drone provided by the present invention, the junction temperature prediction model is obtained by training based on the following method: Obtaining sample operating condition information of the motor system of the sample UAV; Inputting the sample operating condition information, the sample remaining life and the sample environmental parameter into an initial junction temperature prediction model to obtain a predicted junction temperature output by the initial junction temperature prediction model; Based on the predicted junction temperature and the junction temperature label of the sample power device, the model parameters of the initial junction temperature prediction model are iteratively optimized to obtain the junction temperature prediction model.

[0011] According to a method for predicting junction temperature of a power device of a motor controller of a drone provided by the present invention, the method further comprises: Sending the electrical characteristic parameters, the operating condition information and the environmental parameters to a server; receiving an updated first model parameter of the remaining life prediction model and an updated second model parameter of the junction temperature prediction model sent by the server, wherein the first model parameter is a model parameter obtained by the server after iteratively adjusting the remaining life prediction model based on the electrical characteristic parameter and the environmental parameter, and the second model parameter is a model parameter obtained by the server after iteratively adjusting the junction temperature prediction model based on the operating condition information, the environmental parameter and the remaining life output by the remaining life prediction model; The remaining lifetime prediction model is updated based on the first model parameters, and the junction temperature prediction model is updated based on the second model parameters.

[0012] According to a method for predicting junction temperature of a power device of a motor controller of a drone provided by the present invention, the electrical characteristic parameter includes at least one of the following: a saturation voltage drop of the power device, a conduction current of the power device, and a switching frequency of the power device; The operating condition information includes at least one of the following: a motor speed of the motor system, a DC bus voltage of the motor system, a motor phase voltage of the motor system, and a motor phase current of the motor system; The environmental parameters include at least one of the following: the housing temperature of the controller to which the power device belongs, the ambient temperature of the environment in which the drone is located, the ambient air pressure and the ambient wind speed.

[0013] The present invention also provides a junction temperature prediction device for a power device of a UAV motor controller, comprising: An acquisition unit, used to acquire electrical characteristic parameters of a power device of a motor controller of a UAV during operation, operating condition information of a motor system of the UAV, and environmental parameters corresponding to the UAV; A first prediction unit is used to input the electrical characteristic parameters and the environmental parameters into a remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model, wherein the remaining life prediction model is trained based on sample electrical characteristic parameters of sample power devices of sample UAV motor controllers and sample environmental parameters corresponding to the sample UAVs; The second prediction unit is used to input the remaining life, the operating condition information and the environmental parameters into a junction temperature prediction model to obtain the junction temperature of the power device output by the junction temperature prediction model, and the junction temperature prediction model is trained based on the sample remaining life of the sample power device, the sample operating condition information of the motor system of the sample UAV and the sample environmental parameters.

[0014] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, a method for predicting the junction temperature of a power device of a motor controller of a UAV as described above is implemented.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for predicting the junction temperature of a power device of a motor controller of a drone as described in any one of the above is implemented.

[0016] The present invention also provides a computer program product, comprising a computer program, which, when executed by a processor, implements any of the above-mentioned methods for predicting the junction temperature of a power device of a motor controller of a drone.

[0017] The junction temperature prediction method, device and electronic device of the power device of the UAV motor controller provided by the present invention input the electrical characteristic parameters and environmental parameters obtained during the operation of the power device of the UAV motor controller into the remaining life prediction model, obtain the remaining life of the power device output by the remaining life prediction model, and then input the remaining life of the power device, the operating condition information of the motor system of the UAV and the environmental parameters into the junction temperature prediction model, obtain the junction temperature of the power device output by the junction temperature prediction model, thereby realizing the prediction of the junction temperature of the power device. It can be seen that the present invention first inputs the electrical characteristic parameters and environmental parameters of the power device into the remaining life prediction model to predict its remaining life, and then inputs the remaining life, the operating condition information of the motor system and the environmental parameters into the junction temperature prediction model, and then predicts the junction temperature of the power device. This method considers from a multi-dimensional perspective, makes full use of the various parameters of the UAV and the motor controller, and can improve the accuracy and reliability of the prediction of the junction temperature of the power device. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1This is one of the flow charts of the method for predicting the junction temperature of the power device of the motor controller of the UAV provided in the embodiment of the present invention.

[0020] Figure 2 It is a structural schematic diagram of the remaining life prediction model provided by an embodiment of the present invention.

[0021] Figure 3 It is a schematic diagram of the structure of the junction temperature prediction model provided by an embodiment of the present invention.

[0022] Figure 4 It is a training flow chart of the remaining life prediction model provided by an embodiment of the present invention.

[0023] Figure 5 It is a training flow chart of the junction temperature prediction model provided by an embodiment of the present invention.

[0024] Figure 6 This is the second flow chart of the method for predicting the junction temperature of the power device of the motor controller of the UAV provided in the embodiment of the present invention.

[0025] Figure 7 It is a schematic diagram of the overall process of the junction temperature prediction method of the power device of the motor controller of the unmanned aerial vehicle provided by an embodiment of the present invention.

[0026] Figure 8 It is a structural schematic diagram of a junction temperature prediction system for power devices of a motor controller of a UAV provided in an embodiment of the present invention.

[0027] Fig. 9 It is a structural schematic diagram of a junction temperature prediction device for a power device of a motor controller of a UAV provided in an embodiment of the present invention.

[0028] Fig.10 It is a schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] Combine the following Figure 1-Figure 8The present invention describes a method for predicting the junction temperature of a power device of a motor controller of a drone. The execution subject of the method for predicting the junction temperature of a power device of a motor controller of a drone can be an electronic device such as a drone, a terminal or a computer, or a device for predicting the junction temperature of a power device of a motor controller of a drone arranged in the electronic device. The device for predicting the junction temperature of a power device of a motor controller of a drone can be implemented by software, hardware or a combination of the two.

[0031] Figure 1 FIG. 1 is a flow chart of a method for predicting junction temperature of a power device of a motor controller of a UAV provided by an embodiment of the present invention. Figure 1 As shown, the junction temperature prediction method of the UAV motor controller power device includes the following steps: Step 101: Acquire electrical characteristic parameters of a power device of a motor controller of a UAV during operation, operating condition information of a motor system of the UAV, and the environmental parameters.

[0032] The UAV may be a multi-rotor UAV, a compound wing UAV or any type of UAV equipped with a motor drive system, and the motor controller power device may be an insulated gate bipolar transistor (IGBT) or silicon carbide (SiC), etc.

[0033] For example, the electrical characteristic parameter of a power device refers to a parameter obtained by directly measuring the power device and can reflect the degree of aging of the power device. For example, the electrical characteristic parameter includes at least one of the following: the saturation voltage drop of the power device, the on-current of the power device, and the switching frequency of the power device. Environmental parameters may include, for example, the housing temperature of the controller to which the power device belongs. The operating condition information of the motor system refers to the information that can be directly measured by the motor system of the drone. For example, the operating condition information includes at least one of the following: the motor speed of the motor system, the DC bus voltage of the motor system, the motor phase voltage of the motor system, the motor phase current of the motor system, and the average switching frequency of the power device of the motor controller.

[0034] Step 102: input the electrical characteristic parameters and the device environmental parameters into a remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model, wherein the remaining life prediction model is trained based on sample electrical characteristic parameters of sample power devices of sample UAV motor controllers and sample environmental parameters corresponding to the sample UAVs.

[0035] For example, when the electrical characteristic parameters of the power device and the shell temperature of the controller to which the power device belongs are obtained, the electrical characteristic parameters of the power device and the shell temperature of the controller to which the power device belongs are input into a trained remaining life prediction model, and the remaining life prediction model is used to extract and analyze the features of each input parameter, and finally the predicted remaining life of the power device is output. Figure 2 is a schematic diagram of the structure of the remaining life prediction model provided by an embodiment of the present invention, such as Figure 2 As shown, taking the electrical characteristic parameters including the saturation voltage drop and on-current of the power device and the environmental parameters including the shell temperature of the controller to which the power device belongs as an example, the remaining life prediction model finally outputs the remaining life of the power device.

[0036] It should be noted that the remaining life prediction model includes an input layer, a hidden layer and an output layer. Any type of neural network can be used, such as back propagation (BP) neural network, recurrent neural network (RNN), convolutional neural network (CNN), radial basis function network (BFnetwork), etc., and is not limited to a specific type of neural network. Among them, the input layer contains a number of neurons to receive the electrical characteristic parameters of the power device and the environmental parameters such as the shell temperature of the controller to which the power device belongs. The hidden layer includes at least one layer, and each hidden layer includes a number of neurons, which can perform nonlinear transformation and dimensionality reduction processing on the information output by the neurons of the input layer, and then extract feature information. This process is the key to the information transmission and conversion of the neural network. The output layer is responsible for processing the transformed feature information output by the hidden layer, and finally outputs the remaining life of the power device. The remaining life of the power device can describe the characteristics of the power device under the full life cycle range, and can reflect the aging degree of the power device under the current operating state, and try to avoid the influence of parameter drift caused by aging of the power device on the accuracy of junction temperature estimation.

[0037] It should be noted that when the power device is a brand new power device, the maximum life of the power device may also be input into the remaining life prediction model to improve the accuracy of the remaining life prediction of the power device, and the present invention does not limit this.

[0038] Step 103: input the remaining life, the operating condition information and the environmental parameters into a junction temperature prediction model to obtain the junction temperature of the power device output by the junction temperature prediction model, wherein the junction temperature prediction model is trained based on the sample remaining life of the sample power device, the sample operating condition information of the motor system of the sample UAV and the sample environmental parameters.

[0039] The sample remaining life is the remaining life of the sample power device output by the trained remaining life prediction model.

[0040] For example, when the remaining life of a power device output by a remaining life prediction model is obtained, the acquired environmental parameters, the operating condition information of the motor system and the remaining life of the power device output by the remaining life prediction model are input into a junction temperature prediction model, and the junction temperature prediction model is used to extract and analyze the features of each input parameter, and finally the predicted junction temperature of the power device is output. Figure 3 is a schematic diagram of the structure of the junction temperature prediction model provided by an embodiment of the present invention, such as Figure 3 As shown, taking the input data including the remaining life of the power device, the operating condition information including the motor speed, the DC bus voltage and the motor phase current, and the environmental parameters including the shell temperature of the controller to which the power device belongs, the ambient temperature and the ambient air pressure as an example, the junction temperature prediction model finally outputs the junction temperature of the power device.

[0041] It should be noted that the environmental parameters of the input junction temperature prediction model include not only the housing temperature of the controller to which the power device belongs, but also at least one of the following: the ambient temperature, ambient air pressure and ambient wind speed of the environment in which the drone is located, so as to improve the reliability of the junction temperature prediction with multi-dimensional environmental parameters. It should be noted that the junction temperature prediction model includes an input layer, a hidden layer and an output layer, and any type of neural network can be used. The input layer contains a number of neurons for receiving the remaining life of the power device, the operating condition information of the motor system and the environmental information. The hidden layer includes at least one layer, which is responsible for nonlinear processing and dimensionality reduction processing of the output of the input layer. The output layer is responsible for outputting the junction temperature of the power device, thereby effectively completing the accurate estimation of the junction temperature.

[0042] It should be noted that the junction temperature prediction model can also output the power loss of the motor controller power device. The conduction loss and switching loss of the power device during operation are the sources of heat generation. There is a close physical relationship and correlation characteristics between power loss and junction temperature. The features used in the training network are of great significance to both.

[0043] The junction temperature prediction method of the power device of the UAV motor controller provided by the present invention inputs the electrical characteristic parameters and environmental parameters obtained during the operation of the power device of the UAV motor controller into the remaining life prediction model, obtains the remaining life of the power device output by the remaining life prediction model, and then inputs the remaining life of the power device, the operating condition information of the motor system of the UAV and the environmental parameters into the junction temperature prediction model, obtains the junction temperature of the power device output by the junction temperature prediction model, thereby realizing the prediction of the junction temperature of the power device. It can be seen that the present invention first inputs the electrical characteristic parameters and environmental parameters of the power device into the remaining life prediction model to predict its remaining life, and then inputs the remaining life, the operating condition information of the motor system and the environmental parameters into the junction temperature prediction model, and then predicts the junction temperature of the power device. This method considers from a multi-dimensional perspective, makes full use of the various parameters of the UAV and the motor controller, and can improve the accuracy and reliability of the prediction of the junction temperature of the power device.

[0044] In one embodiment, Figure 4 is a training flow chart of the remaining life prediction model provided by an embodiment of the present invention, such as Figure 4 As shown, the remaining life prediction model is trained based on the following method: Step 401: Obtain sample electrical characteristic parameters of sample power devices of the sample UAV motor controller and sample environmental parameters corresponding to the sample UAV.

[0045] For example, multiple sample drones are obtained, and for each sample drone, sample electrical characteristic parameters of a sample power device of a motor controller of the sample drone during operation and the shell temperature of a sample controller to which the sample power device belongs are obtained, and the shell temperature of the sample controller is used as a sample environmental parameter corresponding to the sample drone.

[0046] Step 402: Input the sample electrical characteristic parameters and the sample environmental parameters into an initial remaining life prediction model to obtain a predicted remaining life output by the initial remaining life prediction model.

[0047] For example, the obtained sample electrical characteristic parameters of the sample power device and the shell temperature of the sample controller are input into the initial remaining life prediction model, and the initial remaining life prediction model is used to extract and analyze the input sample parameters, and finally the predicted remaining life of the sample power device is output.

[0048] Step 403: based on the predicted remaining lifetime and the remaining lifetime labels of the sample power devices, iteratively optimize the model parameters of the initial remaining lifetime prediction model to obtain the remaining lifetime prediction model.

[0049] For example, when the predicted remaining life output by the initial remaining life prediction model is obtained, the first loss information is constructed based on the predicted remaining life and the remaining life label of the sample power device, and the model parameters of the initial remaining life prediction model are iteratively optimized based on the first loss information until the convergence conditions are reached, and finally the remaining life prediction model is obtained.

[0050] In this embodiment, the initial remaining life prediction model is trained using sample electrical characteristic parameters of sample power devices of sample UAV motor controllers and sample environmental parameters corresponding to sample UAVs, and based on the predicted remaining life and the remaining life labels of the sample power devices, the model parameters of the initial remaining life prediction model are iteratively optimized to improve the accuracy of the remaining life prediction model finally trained.

[0051] The above step 402 inputs the sample electrical characteristic parameters and the sample environmental parameters into the initial remaining life prediction model to obtain the predicted remaining life output by the initial remaining life prediction model, which can be specifically implemented in the following manner: For each test, the sample electrical characteristic parameters and sample environmental parameters obtained in this test are input into the initial remaining life prediction model to obtain the predicted remaining life of the sample power device corresponding to this test output by the initial remaining life prediction model.

[0052] Accordingly, the above step 403 iteratively optimizes the model parameters of the initial remaining life prediction model based on the predicted remaining life and the remaining life label of the sample power device to obtain the remaining life prediction model, which can be specifically implemented in the following manner: Based on the predicted remaining life of the sample power devices corresponding to each test and the remaining life labels of the sample power devices corresponding to each test, the model parameters of the initial remaining life prediction model are iteratively optimized to obtain the remaining life prediction model.

[0053] For example, the classic BP neural network is used as the type of the two-layer neural network model. The two-layer neural network model refers to the remaining life prediction model and the junction temperature prediction model, and the sigmoid function is used as the nonlinear activation function. The core of the BP neural network is the error back propagation algorithm. During the training process, the error between the output layer data and the target value is evaluated and back propagated to help the neural network update its weights, thereby achieving accurate nonlinear mapping.

[0054] The aging degree and remaining life of power devices can be reflected by the electrical characteristic parameters of power devices when they are working. As the aging degree increases, the electrical characteristic parameters will also change. The number of neurons in the input layer of the remaining life prediction model is 3, corresponding to the three sample electrical characteristic parameters of the saturation voltage drop, on-current, and shell temperature of the controller to which the power device belongs; there are two hidden layers, each containing 7 neurons; the number of neurons in the output layer is 1, which is used to describe the predicted remaining life of the sample power device. The training data of the remaining life prediction model can be obtained by thermodynamic simulation through finite element calculation, or by testing the power device under real conditions. A group of new power devices that have been screened can be used and tested for multiple cycles. The sample electrical characteristic parameters of the sample power devices are collected by configuring the corresponding voltage sampling circuit and current sampling circuit, and the sample environmental parameters of the sample power devices are collected by configuring the corresponding temperature sampling circuit, etc., to obtain the sample electrical characteristic parameters and sample environmental parameters obtained in this test. The sample electrical characteristic parameters and sample environmental parameters obtained by the test are input into the initial remaining life prediction model to obtain the predicted remaining life of the corresponding sample power device output by the initial remaining life prediction model. The same method is used for other tests, so that the predicted remaining life of the sample power device corresponding to each test can be obtained, and then the first loss information is constructed based on the predicted remaining life of the sample power device corresponding to each test and the remaining life label of the sample power device corresponding to each test. The model parameters of the initial remaining life prediction model are iteratively optimized based on the first loss information until the convergence condition is reached to obtain the remaining life prediction model.

[0055] In this embodiment, a data set is established based on sample electrical characteristic parameters, sample environmental parameters and test times of each test of a sample power device. The training data is rich and the characteristic parameters are numerous, which can further improve the accuracy of the remaining life prediction model obtained by training.

[0056] In one embodiment, the remaining life labels of the sample power devices corresponding to each test may be determined in the following manner: The total number of tests corresponding to the upper limit of the service life of the sample power device is obtained; for each test, based on the test sequence number representing the test and the total number of tests, a remaining life label corresponding to the test is determined.

[0057] For example, a group of brand new power devices that have been screened are used to perform multiple cycle tests on the brand new power devices until the power devices reach the upper limit of their service life. When the test starts, the remaining life of the power devices is calibrated as 1, and when the upper limit of the service life is reached, the remaining life of the power devices is calibrated as 0. The remaining life of the power devices corresponding to any number of tests can be expressed by the following formula (1). The remaining life of the power devices corresponding to any number of tests is used as the remaining life label of the corresponding test.

[0058] (1) in, RL Indicates the remaining life of the power device corresponding to the current test; Indicates the test sequence number corresponding to the current test; Indicates the total number of tests corresponding to the upper limit of service life.

[0059] In this embodiment, the remaining life label corresponding to the current test can be determined based on the test sequence number and the total number of tests that characterize the current test, thereby achieving automatic determination of the remaining life label.

[0060] In one embodiment, Figure 5 is a training flow chart of the junction temperature prediction model provided by an embodiment of the present invention, such as Figure 5 As shown, the junction temperature prediction model is trained based on the following method: Step 501: Obtain sample operating condition information of the motor system of the sample UAV.

[0061] For example, the sample operating condition information includes sample motor speed, sample DC bus voltage and sample motor phase current, wherein the sample motor speed will affect the heat dissipation conditions and is an external feature of the system operation. It can be measured by configuring a photoelectric sensor or encoder, or by calculating the speed estimate of a speed sensorless control strategy; the sample DC bus voltage and the sample motor phase current are directly measured by a voltage sensor and a current sensor, respectively.

[0062] Step 502: Input the sample operating condition information, the sample remaining life, and the sample environmental parameters into an initial junction temperature prediction model to obtain a predicted junction temperature output by the initial junction temperature prediction model.

[0063] For example, the initial junction temperature prediction model consists of an input layer, a hidden layer, and an output layer. The input of the input layer of the initial junction temperature prediction model includes sample operating condition information of the motor system, sample environmental parameters, and sample remaining life of the sample power device output by the trained remaining life prediction neural network. The sample environmental information includes the shell temperature of the sample controller to which the sample power device belongs, the sample environmental temperature, and the sample environmental air pressure. Therefore, the input layer contains a total of 7 neurons, corresponding to the seven input information. There are two hidden layers, each containing 15 neurons, which perform nonlinear processing and dimensionality reduction on the input data. The output layer has a total of 2 neurons, which output the predicted junction temperature of the sample power device. The shell temperature of the sample controller can be measured by an embedded temperature sensor; the sample environmental temperature can be obtained by the temperature and humidity sensor of the sample drone; the sample environmental air pressure characterizes the air density of the sample environment in which the sample drone is located, directly affects the propeller thrust output characteristics of the sample drone, the heat dissipation conditions of the sample controller, etc., and can be obtained by measuring the altimeter of the sample drone.

[0064] Step 503: Based on the predicted junction temperature and the junction temperature label of the sample power device, iteratively optimize the model parameters of the initial junction temperature prediction model to obtain the junction temperature prediction model.

[0065] For example, when the predicted junction temperature output by the initial junction temperature prediction model is obtained, the second loss information is constructed based on the predicted junction temperature and the junction temperature label of the sample power device, and the model parameters of the initial junction temperature prediction model are iteratively optimized based on the second loss information until the convergence conditions are reached, and finally the junction temperature prediction model is obtained.

[0066] In this embodiment, the initial junction temperature prediction model is trained using sample operating condition information of a sample UAV, sample remaining life of a sample power device, and sample environmental parameters, and based on the predicted junction temperature and the junction temperature label of the sample power device, the model parameters of the initial junction temperature prediction model are iteratively optimized to improve the accuracy of the junction temperature prediction model finally trained.

[0067] In one embodiment, Figure 6 FIG. 2 is a flow chart of a method for predicting junction temperature of a power device of a motor controller of a UAV provided by an embodiment of the present invention. Figure 6 As shown, the method for predicting the junction temperature of the power device of the UAV motor controller also includes the following steps: Step 601: Send the electrical characteristic parameters, the operating condition information and the environmental parameters to a server.

[0068] For example, the effective training of the two-layer neural network model relies on a large amount of computing resources, and the process is complicated and time-consuming. The UAV motor controller usually uses an embedded processor as the main control unit, and the computing power is extremely limited, which makes it difficult to meet the needs of fast and effective training of the two-layer neural network model locally. In addition, the local storage space of the UAV motor controller is also limited, and the real data generated in the actual operation of the UAV cannot be effectively utilized, but only the data under experimental conditions can be used for network training, resulting in a waste of resources. Therefore, the junction temperature prediction method of the present invention establishes a UAV-server collaborative platform, and uses the server to store the newly generated training data of the UAV, that is, the UAV sends the electrical characteristic parameters of the power device generated during operation (saturation voltage drop of the power device, conduction current), the operating condition information of the motor system (DC bus voltage of the motor system, motor phase voltage of the motor system and motor phase current of the motor system) and environmental parameters (shell temperature of the controller to which the power device belongs, ambient temperature, ambient air pressure and ambient wind speed of the environment in which the UAV is located) to the server regularly to avoid local storage of training data. After preprocessing the received training data, the server uses the server's abundant computing resources to retrain the two-layer neural network model, and regularly pushes the updated model parameters obtained after retraining to the drone motor controller, so that the controller completes the update of the corresponding model based on the received updated model parameters.

[0069] Step 602: Receive the updated first model parameters of the remaining life prediction model and the updated second model parameters of the junction temperature prediction model sent by the server, wherein the first model parameters are model parameters obtained by the server after iteratively adjusting the remaining life prediction model based on the electrical characteristic parameters and the environmental parameters, and the second model parameters are model parameters obtained by the server after iteratively adjusting the junction temperature prediction model based on the operating condition information, the environmental parameters and the remaining life output by the remaining life prediction model.

[0070] For example, when the server receives the training data sent by the drone, it first evaluates the similarity between the received training data and the stored training data. When it is determined that the similarity between the received training data and the stored training data is greater than the preset similarity, it means that the received training data is repeated and has low value, and can be excluded from the training set; when the similarity between the received training data and the stored training data is less than or equal to the preset similarity, it means that the received training data differs greatly and has high value, and the received training data can be added to the training set. Then, the remaining life prediction model is adjusted based on the new training set to obtain the updated first model parameters; and based on the operating condition information, environmental parameters and the remaining life output by the remaining life prediction model in the new training set, the junction temperature prediction model is fine-tuned to obtain the updated second model parameters. The updated first model parameters and the updated second model parameters are sent to the drone.

[0071] It should be noted that the conditions for the end of adjustment need to meet the following conditions: the difference between the remaining life corresponding to the last adjustment and the remaining life corresponding to this adjustment is greater than the preset value, or the number of network training times is greater than the preset number. If either of these two conditions is not met, continue to update the model parameters.

[0072] Step 603: updating the remaining lifetime prediction model based on the first model parameters, and updating the junction temperature prediction model based on the second model parameters.

[0073] For example, an Over-The-Air (OTA) function may be used to update the model parameters of the remaining lifetime prediction model based on the first model parameters, and to update the junction temperature prediction model based on the second model parameters.

[0074] It should be noted that the execution entities of the above steps 401 to 403 and the execution entities of the above steps 501 to 503 can both be servers, that is, the server can complete the initial training of the remaining life prediction model and the junction temperature prediction model, and can also complete the update of the remaining life prediction model and the junction temperature prediction model.

[0075] In this embodiment, the server uses the electrical characteristic parameters of the power devices, the operating condition information of the motor system and the environmental parameters generated by the drone to fine-tune the model parameters of the remaining life prediction model and the junction temperature prediction model, thereby further improving the accuracy of the remaining life prediction model and the junction temperature prediction model.

[0076] In one embodiment, the electrical characteristic parameter is at least one of the following: a saturation voltage drop of the power device, a conduction current of the power device, and a switching frequency of the power device; The operating condition information includes at least one of the following: a motor speed of the motor system, a DC bus voltage of the motor system, a motor phase voltage of the motor system, and a motor phase current of the motor system; The environmental parameters include at least one of the following: the housing temperature of the controller to which the power device belongs, the ambient temperature of the environment in which the drone is located, the ambient air pressure and the ambient wind speed.

[0077] Figure 7 FIG. 1 is a schematic diagram of the overall process of the method for predicting the junction temperature of the power device of the UAV motor controller provided by an embodiment of the present invention. Figure 7 As shown, the electrical characteristic parameters of the power device of the UAV motor controller during operation and the environmental parameters corresponding to the UAV are input into the remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model; then the operating condition information, environmental parameters and the remaining life of the power device output by the remaining life prediction model of the UAV motor system are input into the junction temperature prediction model to obtain the junction temperature and power loss information of the power device output by the junction temperature prediction model. In addition, the UAV sends the electrical characteristic parameters of the power device of the UAV motor controller during operation, the operating condition information and environmental parameters of the UAV motor system to the server as training data, and the server performs similarity calculation based on the stored training data and the received training data, selects valuable training data and adds them to the training set, and then updates the model parameters of the remaining life prediction model and the junction temperature prediction model based on the new training set, respectively, to obtain the updated first model parameters and the updated second model parameters, and sends the updated first model parameters and the updated second model parameters to the UAV, so that the UAV completes the update of the remaining life prediction model and the junction temperature prediction model based on the received updated first model parameters and the updated second model parameters.

[0078] Figure 8 FIG. 1 is a schematic diagram of the structure of a junction temperature prediction system for a power device of a UAV motor controller provided by an embodiment of the present invention. Figure 8As shown, the junction temperature prediction system of the power device of the UAV motor controller includes a UAV and a server, and the UAV includes a data acquisition module, a remaining life prediction model and a junction temperature prediction model; wherein the data acquisition module is configured to obtain relevant data such as electrical characteristic parameters of the power device, operating condition information of the motor system and environmental parameters, and is mainly composed of various sensors such as voltage sensors, current sensors, temperature sensors, altimeters and their related circuits; the remaining life prediction model is configured to calculate the remaining life of the power device using the collected data, and estimate and guide its aging degree; the junction temperature prediction model is configured to calculate the power loss information and junction temperature of the power device using the collected data, and realize accurate estimation of the junction temperature; the server communicates with the UAV, and the UAV uploads the real-time operation data to the server as training data, and the server adjusts the model parameters of the remaining life prediction model and the junction temperature prediction model based on the received training data, and of course, the initial training and deployment of the remaining life prediction model and the junction temperature prediction model can also be completed, and the communication between the server and the UAV can be wired or wireless, which is not limited by the present invention.

[0079] The junction temperature prediction system of the power device of the motor controller of the unmanned aerial vehicle provided by the embodiment of the present invention is used for the accurate junction temperature estimation of the power device in the unmanned aerial vehicle. By establishing a double-layer neural network model, the remaining service life and junction temperature of the power device are accurately and quickly estimated. In combination with the server, it is possible to realize the timed update of the double-layer neural network model without occupying the local computing resources of the unmanned aerial vehicle motor controller, which can effectively improve the reliability of the double-layer neural network model. In addition, the embodiment of the present invention fully considers the operating condition information of the motor system during the flight of the unmanned aerial vehicle, and uses the operating condition information of the motor system that is easily obtained as the network input parameter. Overall, the junction temperature estimation of the power device can be realized, the actual working state of the controller is fully considered, and the accuracy of the junction temperature estimation under the whole life cycle is ensured. It can serve as an important reference for the working state of the unmanned aerial vehicle, and can also guide the formulation of the flight control strategy, effectively improving the reliability of the operation of the unmanned aerial vehicle system.

[0080] The junction temperature prediction device of the UAV motor controller power device provided by the present invention is described below. The junction temperature prediction device of the UAV motor controller power device described below and the junction temperature prediction method of the UAV motor controller power device described above can be referenced to each other.

[0081] Fig. 9 is a schematic diagram of the structure of a junction temperature prediction device for a power device of a UAV motor controller provided by an embodiment of the present invention, such as Fig. 9 As shown, the junction temperature prediction device 900 of the power device of the UAV motor controller includes an acquisition unit 901, a first prediction unit 902 and a second prediction unit 903; wherein: An acquisition unit 901 is used to acquire electrical characteristic parameters of a power device of a motor controller of a UAV during operation, operating condition information of a motor system of the UAV, and environmental parameters corresponding to the UAV; A first prediction unit 902 is used to input the electrical characteristic parameters and the environmental parameters into a remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model, wherein the remaining life prediction model is trained based on sample electrical characteristic parameters of sample power devices of sample UAV motor controllers and sample environmental parameters corresponding to the sample UAVs; The second prediction unit 903 is used to input the remaining life, the operating condition information and the environmental parameters into a junction temperature prediction model to obtain the junction temperature of the power device output by the junction temperature prediction model. The junction temperature prediction model is trained based on the sample remaining life of the sample power device, the sample operating condition information of the motor system of the sample UAV and the sample environmental parameters.

[0082] The junction temperature prediction device of the power device of the UAV motor controller provided by the present invention inputs the electrical characteristic parameters and environmental parameters obtained during the operation of the power device of the UAV motor controller into the remaining life prediction model, obtains the remaining life of the power device output by the remaining life prediction model, and then inputs the remaining life of the power device, the operating condition information of the motor system of the UAV and the environmental parameters into the junction temperature prediction model, obtains the junction temperature of the power device output by the junction temperature prediction model, thereby realizing the prediction of the junction temperature of the power device. It can be seen that the present invention first inputs the electrical characteristic parameters and environmental parameters of the power device into the remaining life prediction model to predict its remaining life, and then inputs the remaining life, the operating condition information of the motor system and the environmental parameters into the junction temperature prediction model, and then predicts the junction temperature of the power device. This method considers from a multi-dimensional perspective, makes full use of the various parameters of the UAV and the motor controller, and can improve the accuracy and reliability of the prediction of the junction temperature of the power device.

[0083] Based on any of the above embodiments, the remaining life prediction model is trained based on the following method: Obtaining sample electrical characteristic parameters of sample power devices of the sample UAV motor controller and sample environmental parameters corresponding to the sample UAV; Inputting the sample electrical characteristic parameters and the sample environmental parameters into an initial remaining life prediction model to obtain a predicted remaining life output by the initial remaining life prediction model; Based on the predicted remaining life and the remaining life labels of the sample power devices, model parameters of the initial remaining life prediction model are iteratively optimized to obtain the remaining life prediction model.

[0084] Based on any of the above embodiments, the step of inputting the sample electrical characteristic parameters and the sample environmental parameters into an initial remaining life prediction model to obtain a predicted remaining life output by the initial remaining life prediction model includes: For each test, the sample electrical characteristic parameters and sample environmental parameters obtained in this test are input into the initial remaining life prediction model to obtain the predicted remaining life of the sample power device corresponding to this test output by the initial remaining life prediction model; The iterative optimization of the model parameters of the initial remaining life prediction model based on the predicted remaining life and the remaining life label of the sample power device to obtain the remaining life prediction model includes: Based on the predicted remaining life of the sample power devices corresponding to each test and the remaining life labels of the sample power devices corresponding to each test, the model parameters of the initial remaining life prediction model are iteratively optimized to obtain the remaining life prediction model.

[0085] Based on any of the above embodiments, the junction temperature prediction device 900 of the UAV motor controller power device further includes: Obtaining a total number of tests corresponding to an upper limit of the service life of the sample power device; For each test, based on the test sequence number representing the test and the total number of tests, the remaining life label corresponding to the test is determined.

[0086] Based on any of the above embodiments, the junction temperature prediction model is trained based on the following method: Obtaining sample operating condition information of the motor system of the sample UAV; Inputting the sample operating condition information, the sample remaining life and the sample environmental parameter into an initial junction temperature prediction model to obtain a predicted junction temperature output by the initial junction temperature prediction model; Based on the predicted junction temperature and the junction temperature label of the sample power device, the model parameters of the initial junction temperature prediction model are iteratively optimized to obtain the junction temperature prediction model.

[0087] Based on any of the above embodiments, the junction temperature prediction device 900 of the UAV motor controller power device further includes: A sending unit, used for sending the electrical characteristic parameter, the operating condition information and the environmental parameter to a server; a receiving unit, configured to receive an updated first model parameter of the remaining life prediction model and an updated second model parameter of the junction temperature prediction model sent by the server, wherein the first model parameter is a model parameter obtained by the server after iteratively adjusting the remaining life prediction model based on the electrical characteristic parameter and the environmental parameter, and the second model parameter is a model parameter obtained by the server after iteratively adjusting the junction temperature prediction model based on the operating condition information, the environmental parameter and the remaining life output by the remaining life prediction model; An updating unit is used to update the remaining life prediction model based on the first model parameters, and to update the junction temperature prediction model based on the second model parameters.

[0088] Based on any of the above embodiments, the electrical characteristic parameter is at least one of the following: a saturation voltage drop of the power device, a conduction current of the power device, and a switching frequency of the power device; The operating condition information includes at least one of the following: a motor speed of the motor system, a DC bus voltage of the motor system, a motor phase voltage of the motor system, and a motor phase current of the motor system; The environmental parameters include at least one of the following: the housing temperature of the controller to which the power device belongs, the ambient temperature, ambient air pressure and ambient wind speed of the environment in which the drone is located.

[0089] Fig.10 is a schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention, such as Fig.10As shown, the electronic device may include: a processor (processor) 1010 , a communication interface (Communications Interface) 1020 , a memory (memory) 1030 and a communication bus 1040 , wherein the processor 1010 , the communication interface 1020 , and the memory 1030 communicate with each other via the communication bus 1040 . The processor 1010 can call the logic instructions in the memory 1030 to execute the junction temperature prediction method of the power device of the drone motor controller, the method including: obtaining the electrical characteristic parameters of the power device of the drone motor controller during operation, the operating condition information of the motor system of the drone and the environmental parameters corresponding to the drone; inputting the electrical characteristic parameters and the environmental parameters into the remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model, wherein the remaining life prediction model is trained based on the sample electrical characteristic parameters of the sample power device of the sample drone motor controller and the sample environmental parameters corresponding to the sample drone; inputting the remaining life, the operating condition information and the environmental parameters into the junction temperature prediction model to obtain the junction temperature of the power device output by the junction temperature prediction model, wherein the junction temperature prediction model is trained based on the sample remaining life of the sample power device, the sample operating condition information of the motor system of the sample drone and the sample environmental parameters.

[0090] In addition, the logic instructions in the above-mentioned memory 1030 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0091] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the junction temperature prediction method of the power device of the drone motor controller provided by the above methods, and the method includes: obtaining the electrical characteristic parameters of the power device of the drone motor controller during operation, the operating condition information of the motor system of the drone, and the environmental parameters corresponding to the drone; inputting the electrical characteristic parameters and the environmental parameters into a remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model, and the remaining life prediction model is trained based on the sample electrical characteristic parameters of the sample power device of the sample drone motor controller and the sample environmental parameters corresponding to the sample drone; inputting the remaining life, the operating condition information and the environmental parameters into the junction temperature prediction model to obtain the junction temperature of the power device output by the junction temperature prediction model, and the junction temperature prediction model is trained based on the sample remaining life of the sample power device, the sample operating condition information of the motor system of the sample drone, and the sample environmental parameters.

[0092] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the junction temperature prediction method of the power device of the drone motor controller provided by the above-mentioned methods, the method comprising: obtaining electrical characteristic parameters of the power device of the drone motor controller during operation, operating condition information of the motor system of the drone, and environmental parameters corresponding to the drone; inputting the electrical characteristic parameters and the environmental parameters into a remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model, wherein the remaining life prediction model is trained based on sample electrical characteristic parameters of sample power devices of sample drone motor controllers and sample environmental parameters corresponding to the sample drone; inputting the remaining life, the operating condition information, and the environmental parameters into a junction temperature prediction model to obtain the junction temperature of the power device output by the junction temperature prediction model, wherein the junction temperature prediction model is trained based on the sample remaining life of the sample power device, the sample operating condition information of the motor system of the sample drone, and the sample environmental parameters.

[0093] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0094] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting junction temperature of power devices of a UAV motor controller, characterized in that: include: Obtaining electrical characteristic parameters of a power device of a UAV motor controller during operation, operating condition information of a motor system of the UAV, and environmental parameters corresponding to the UAV; Inputting the electrical characteristic parameters and the environmental parameters into a remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model, wherein the remaining life prediction model is trained based on sample electrical characteristic parameters of sample power devices of sample UAV motor controllers and sample environmental parameters corresponding to the sample UAVs; The remaining life, the operating condition information and the environmental parameters are input into a junction temperature prediction model to obtain the junction temperature of the power device output by the junction temperature prediction model. The junction temperature prediction model is trained based on the sample remaining life of the sample power device, the sample operating condition information of the motor system of the sample UAV and the sample environmental parameters.

2. The method for predicting junction temperature of power devices of a UAV motor controller according to claim 1, characterized in that: The remaining life prediction model is trained based on the following method: Obtaining sample electrical characteristic parameters of sample power devices of the sample UAV motor controller and sample environmental parameters corresponding to the sample UAV; Inputting the sample electrical characteristic parameters and the sample environmental parameters into an initial remaining life prediction model to obtain a predicted remaining life output by the initial remaining life prediction model; Based on the predicted remaining life and the remaining life labels of the sample power devices, model parameters of the initial remaining life prediction model are iteratively optimized to obtain the remaining life prediction model.

3. The method for predicting junction temperature of power devices of a UAV motor controller according to claim 2, characterized in that: The step of inputting the sample electrical characteristic parameters and the sample environmental parameters into an initial remaining life prediction model to obtain a predicted remaining life output by the initial remaining life prediction model comprises: For each test, the sample electrical characteristic parameters and sample environmental parameters obtained in this test are input into the initial remaining life prediction model to obtain the predicted remaining life of the sample power device corresponding to this test output by the initial remaining life prediction model; The iterative optimization of the model parameters of the initial remaining life prediction model based on the predicted remaining life and the remaining life label of the sample power device to obtain the remaining life prediction model includes: Based on the predicted remaining life of the sample power devices corresponding to each test and the remaining life labels of the sample power devices corresponding to each test, the model parameters of the initial remaining life prediction model are iteratively optimized to obtain the remaining life prediction model.

4. The method for predicting junction temperature of power devices of a UAV motor controller according to claim 3, characterized in that: The method further comprises: Obtaining a total number of tests corresponding to an upper limit of the service life of the sample power device; For each test, based on the test sequence number representing the test and the total number of tests, the remaining life label corresponding to the test is determined.

5. The method for predicting junction temperature of power devices of a UAV motor controller according to claim 1, characterized in that: The junction temperature prediction model is trained based on the following method: Obtaining sample operating condition information of the motor system of the sample UAV; Inputting the sample operating condition information, the sample remaining life and the sample environmental parameter into an initial junction temperature prediction model to obtain a predicted junction temperature output by the initial junction temperature prediction model; Based on the predicted junction temperature and the junction temperature label of the sample power device, the model parameters of the initial junction temperature prediction model are iteratively optimized to obtain the junction temperature prediction model.

6. The method for predicting junction temperature of a power device of a UAV motor controller according to any one of claims 1 to 5, characterized in that: The method further comprises: Sending the electrical characteristic parameters, the operating condition information and the environmental parameters to a server; receiving an updated first model parameter of the remaining life prediction model and an updated second model parameter of the junction temperature prediction model sent by the server, wherein the first model parameter is a model parameter obtained by the server after iteratively adjusting the remaining life prediction model based on the electrical characteristic parameter and the environmental parameter, and the second model parameter is a model parameter obtained by the server after iteratively adjusting the junction temperature prediction model based on the operating condition information, the environmental parameter and the remaining life output by the remaining life prediction model; The remaining lifetime prediction model is updated based on the first model parameters, and the junction temperature prediction model is updated based on the second model parameters.

7. The method for predicting junction temperature of a power device of a UAV motor controller according to any one of claims 1 to 5, characterized in that: The electrical characteristic parameter includes at least one of the following: a saturation voltage drop of the power device, a conduction current of the power device, and a switching frequency of the power device; The operating condition information includes at least one of the following: a motor speed of the motor system, a DC bus voltage of the motor system, a motor phase voltage of the motor system, and a motor phase current of the motor system; The environmental parameters include at least one of the following: the housing temperature of the controller to which the power device belongs, the ambient temperature, ambient air pressure and ambient wind speed of the environment in which the drone is located.

8. A junction temperature prediction device for a power device of a drone motor controller, characterized in that: include: An acquisition unit, used to acquire electrical characteristic parameters of a power device of a motor controller of a UAV during operation, operating condition information of a motor system of the UAV, and environmental parameters corresponding to the UAV; A first prediction unit is used to input the electrical characteristic parameters and the environmental parameters into a remaining life prediction model to obtain the remaining life of the power device output by the remaining life prediction model, wherein the remaining life prediction model is trained based on sample electrical characteristic parameters of sample power devices of sample UAV motor controllers and sample environmental parameters corresponding to the sample UAVs; The second prediction unit is used to input the remaining life, the operating condition information and the environmental parameters into a junction temperature prediction model to obtain the junction temperature of the power device output by the junction temperature prediction model, and the junction temperature prediction model is trained based on the sample remaining life of the sample power device, the sample operating condition information of the motor system of the sample UAV and the sample environmental parameters.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for predicting the junction temperature of the power device of the motor controller of the unmanned aerial vehicle according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting the junction temperature of a power device of a motor controller for a drone as claimed in any one of claims 1 to 7 is implemented.

Images