A method and apparatus for early warning of faults in a motor drive

Abstract

A fault early warning method and device of a motor driver are used for realizing fault early warning of the motor driver. Embodiments of the present application provide a fault early warning method for a motor driver, comprising: acquiring running state data of the motor driver when the motor driver runs; inputting the running state data into a preset artificial intelligence (AI) diagnosis model, outputting a first abnormality degree parameter of the motor driver through the AI diagnosis model, wherein the AI diagnosis model is used for extracting data features according to the running state data and performing decision reasoning through the data features; and performing fault early warning on the motor driver according to the first abnormality degree parameter.

Description

Technical Field

[0001] This application relates to the field of power electronic power conversion, and more particularly to a fault warning method and apparatus for motor drives. Background Technology

[0002] Motor drivers are used for driving, idling, and braking control of motors. Motor drivers may include power semiconductor modules, which are core components of electric vehicle powertrains and are also prone to failure.

[0003] Current fault diagnosis solutions for motor drives mainly involve monitoring parameters such as the motor drive's temperature, temperature amplitude changes, and cycle number, and then using a life model to assess the motor drive's lifespan. This life model can include the Coffin-Manson model, the Norris-Landzberg model, and the Bayer model.

[0004] In the aforementioned fault diagnosis scheme, the input parameters of the lifespan model are obtained by conducting lifespan tests on a small number of motor drives. This lifespan model can only represent the overall situation in a statistical sense and has no practical value for each individual motor drive. Therefore, the aforementioned lifespan model has poor versatility and cannot be used for fault diagnosis of a single motor drive. Summary of the Invention

[0005] This application provides a fault warning method and apparatus for motor drivers, which are used to implement fault warning for motor drivers.

[0006] To address the aforementioned technical problems, this application provides the following technical solutions:

[0007] In a first aspect, embodiments of this application provide a fault early warning method for a motor driver, comprising:

[0008] Obtain the running state data of the motor driver during operation;

[0009] The operating state data is input into a preset artificial intelligence (AI) diagnostic model, and the AI ​​diagnostic model outputs the first anomaly parameter of the motor driver. The AI ​​diagnostic model is used to extract data features based on the operating state data and to make decision-making inferences based on the data features.

[0010] The motor driver is given a fault warning based on the first anomaly parameter.

[0011] In this solution, since the vehicle is in operation when the motor driver is running in this embodiment of the application, the motor driver can be given a fault warning based on the running status data of the motor driver. The running status data of the motor driver is the data collected when the motor driver is running. The running status data can reflect the real running status of the motor driver. Based on the running status data, the fault warning can be given in advance to predict the failure of the motor driver and realize the fault warning of the motor driver.

[0012] In one possible implementation, the method further includes:

[0013] Static data obtained by low-frequency sampling of the motor driver when the vehicle is stationary, wherein the motor driver is installed on the vehicle;

[0014] The static data is input into the AI ​​diagnostic model, and the AI ​​diagnostic model outputs the second anomaly parameter of the motor driver.

[0015] The motor driver is given a fault warning based on the second anomaly parameter.

[0016] In this solution, static data of the motor driver is acquired when the vehicle is stationary. Based on this static data, a fault warning can be given to the motor driver. The static data of the motor driver is the data collected from the motor driver when the vehicle is stationary. The static data can reflect the true stationary state of the motor driver. Based on this static data, a fault warning can be given, which can realize the early prediction of motor driver failure and achieve fault warning of the motor driver.

[0017] In one possible implementation, the vehicle is in a stopped phase, including at least one of the following:

[0018] The vehicle is in the charging stage, the vehicle is in the starting stage, and the vehicle is in the off stage.

[0019] In this scheme, static data is sampled during the vehicle charging phase, vehicle ignition phase, and vehicle shutdown phase, so that the fault diagnosis device can provide emergency response function in the event of network outage based on the static data.

[0020] In one possible implementation, the step of providing fault warning for the motor driver based on the first anomaly parameter includes:

[0021] The AI ​​diagnostic model dynamically updates the preset abnormal conditions to obtain the updated abnormal conditions.

[0022] When the first anomaly parameter meets the updated anomaly condition, a fault warning is issued for the motor driver.

[0023] In this scheme, abnormal conditions can be preset, and these abnormal conditions can be dynamically updated through an AI diagnostic model to ensure that they can be used for fault diagnosis and to guarantee the accuracy of fault diagnosis. After dynamically updating the abnormal conditions, it can be determined whether the first abnormality degree meets the dynamically updated abnormal conditions. When the first abnormality degree parameter meets the updated abnormal conditions, it indicates that the motor driver may have a fault, and a fault warning is issued to the motor driver. In this embodiment, by dynamically updating the abnormal conditions, the accuracy of fault warning can be further improved.

[0024] In one possible implementation, after acquiring the runtime data of the motor driver, the method further includes:

[0025] The running data is cleaned to obtain cleaned running data;

[0026] The step of inputting the runtime data into a preset artificial intelligence (AI) diagnostic model includes:

[0027] The cleaned and processed operational data is input into the AI ​​diagnostic model.

[0028] In this approach, after acquiring multiple sets of operational data, the data can be cleaned first. For example, data that meets cleaning criteria can be removed. These criteria might include incomplete data or a metric exceeding a certain limit. After obtaining the cleaned operational data, it is input into an AI diagnostic model. The AI ​​diagnostic model extracts data features from the cleaned data. By cleaning the operational data before inputting it into the AI ​​diagnostic model, the inference efficiency of the AI ​​diagnostic model can be improved.

[0029] In one possible implementation, the method further includes:

[0030] The cloud server acquires training sample data from multiple vehicles of the same type and multiple vehicles of different types, and the motor driver is installed on the vehicle;

[0031] The cloud server uses training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types to train a model based on a fault warning task, so as to obtain the preset AI diagnostic model.

[0032] In this solution, the cloud server uses multiple training sample data, which can include training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types. The resulting AI diagnostic model is a global AI model, which is designed for specific fault warning tasks. This model is trained based on all collected vehicle terminal device data and provides end-side inference results for each vehicle terminal device based on this model.

[0033] In one possible implementation, the method further includes:

[0034] The cloud server acquires training sample data for one vehicle or training sample data for multiple vehicles of the same type, and the motor driver is installed on the vehicle.

[0035] The cloud server uses training sample data from one vehicle or training sample data from multiple vehicles of the same type to train a model based on a fault warning task, so as to obtain the preset AI diagnostic model.

[0036] In this solution, the cloud server uses multiple training sample data, which may include training sample data of a single vehicle or training sample data of multiple vehicles of the same type. The resulting AI diagnostic model is an AI local model, which is designed for specific fault warning tasks. This AI local model may be a specific model of vehicle terminal device or even a specific vehicle terminal device. Based on this model, each edge-side inference result is given.

[0037] In one possible implementation, when the cloud server outputs a first anomaly parameter of the motor driver through the AI ​​diagnostic model, the step of providing a fault warning for the motor driver based on the first anomaly parameter includes:

[0038] The cloud server generates fault warning information based on the first anomaly parameter.

[0039] The cloud server sends the fault warning information to the vehicle terminal equipment.

[0040] In this solution, the cloud server in this application embodiment has model training capability and decision reasoning capability. The vehicle terminal device can perform fault warning according to the decision result of the cloud server. Through the interaction between the cloud server and the vehicle terminal device, fault warning of the motor drive can be realized.

[0041] In one possible implementation, when the cloud server sends the AI ​​diagnostic model to the vehicle terminal device, the method further includes: the vehicle terminal device receiving the AI ​​diagnostic model sent by the cloud server.

[0042] The step of inputting the operational data into a preset artificial intelligence (AI) diagnostic model and outputting the first anomaly parameter of the motor driver through the AI ​​diagnostic model includes:

[0043] The vehicle terminal device inputs the operating status data into the AI ​​diagnostic model, and outputs the first anomaly parameter of the motor driver through the AI ​​diagnostic model.

[0044] In this solution, the cloud server in this application embodiment has model training capability, and the vehicle terminal device has decision reasoning capability. The vehicle terminal device can receive the AI ​​diagnostic model from the cloud server, so that the vehicle terminal device can output decision results according to the AI ​​diagnostic model and provide fault warning for the motor drive. Through the interaction between the cloud server and the vehicle terminal device, fault warning for the motor drive can be realized.

[0045] In one possible implementation, the method further includes:

[0046] The cloud server acquires training sample data from multiple different types of vehicles, or training sample data from a single vehicle, or training sample data from multiple vehicles of the same type.

[0047] The cloud server extracts data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations.

[0048] The cloud server sends the pre-trained model to the vehicle terminal device.

[0049] In this solution, the cloud server does not need to train the AI ​​diagnostic model. Instead, the cloud server only needs to extract the data correlations, obtain the pre-trained model, and then send the pre-trained model to the vehicle terminal device.

[0050] In one possible implementation, the method further includes:

[0051] The vehicle terminal device receives a pre-trained model from the cloud server;

[0052] The vehicle terminal device performs model parameter optimization on the pre-trained model based on preset label data to obtain the preset AI diagnostic model.

[0053] In this scheme, the vehicle terminal device performs model parameter optimization on the pre-trained model based on preset label data. Model parameter optimization can also be called fine-tuning, that is, the vehicle terminal device can train the model to obtain the preset AI diagnostic model.

[0054] In one possible implementation, the runtime data includes: dynamic data collected by default during the operation of the motor driver; or,

[0055] The operational data includes dynamic data collected via the CAN bus of the controller area network.

[0056] In one possible implementation, the operational data includes dynamic data obtained by high-frequency sampling when the vehicle is in operation, and the motor driver is mounted on the vehicle.

[0057] In this scheme, the vehicle has multiple states, such as being in an operating phase, which refers to the stage after the vehicle is started, or being in motion. The dynamic data obtained by high-frequency sampling when the vehicle is in the operating phase can be the aforementioned operating state data. Subsequent embodiments will provide examples of the high-frequency sampling data items and data volume. High-frequency sampling refers to data acquisition at a frequency greater than a preset frequency threshold.

[0058] In one possible implementation, the operational data includes at least one of the following: parameters of the motor driver, parameters of the cooling system, parameters of the motor, and parameters of the battery.

[0059] In this solution, the running state data of the motor driver in this application embodiment can be implemented in multiple ways. For example, when the motor driver is running, data can be collected on the motor driver to obtain the parameters of the motor driver, or data can be collected on the cooling system to obtain the parameters of the cooling system, or data can be collected on the motor to obtain the parameters of the motor, or data can be collected on the battery to obtain the parameters of the battery.

[0060] In one possible implementation, the parameters of the motor driver include at least one of the following: DC bus voltage, DC bus current, low-voltage power supply voltage, vehicle driving status command, vehicle status, motor driver temperature, rotor position electrical angle, effective value of motor driver three-phase current, motor driver three-phase current, sampled value of motor driver three-phase current, and motor driver derating status; and / or,

[0061] The parameters of the cooling system include at least one of the following: coolant flow rate, coolant temperature, oil pump target speed, oil pump status, oil pump actual speed, oil pump power supply voltage, oil pump power semiconductor device temperature, and oil pump current; and / or,

[0062] The parameters of the motor include at least one of the following: electromagnetic frequency, motor operating mode command, motor controller operating status, motor target torque command, motor current torque, motor target speed command, motor current speed, motor temperature, motor direct-axis voltage, motor direct-axis current setpoint, motor direct-axis current feedback value, motor quadrature-axis voltage, motor quadrature-axis current setpoint, motor quadrature-axis current feedback value, and motor current verification torque; and / or,

[0063] The parameters of the battery include at least one of the following: rated output voltage, battery capacity, and maximum output current.

[0064] Secondly, embodiments of this application also provide a fault warning device for a motor driver, characterized in that it includes:

[0065] The acquisition module is used to acquire the running state data of the motor driver during operation;

[0066] The reasoning module is used to input the running state data into a preset artificial intelligence (AI) diagnostic model, and output the first anomaly parameter of the motor driver through the AI ​​diagnostic model. The AI ​​diagnostic model is used to extract data features based on the running state data and to perform decision reasoning based on the data features.

[0067] The early warning module is used to provide fault warnings to the motor driver based on the first anomaly parameter.

[0068] In one possible implementation, the acquisition module is further configured to acquire static data obtained by low-frequency sampling of the motor driver when the vehicle is in a stopped state, wherein the motor driver is installed on the vehicle;

[0069] The inference module is also used to input the static data into the AI ​​diagnostic model and output the second anomaly parameter of the motor driver through the AI ​​diagnostic model;

[0070] The early warning module is also used to provide fault warnings for the motor driver based on the second anomaly parameter.

[0071] In one possible implementation, the vehicle is in a stopped phase, including at least one of the following:

[0072] The vehicle is in the charging stage, the vehicle is in the starting stage, and the vehicle is in the off stage.

[0073] In one possible implementation, the early warning module is used to dynamically update the preset abnormal conditions through the AI ​​diagnostic model to obtain the updated abnormal conditions; when the first abnormality parameter meets the updated abnormal conditions, a fault warning is issued to the motor driver.

[0074] In one possible implementation, the device further includes: a data processing module, used to clean the running state data of the motor driver after the acquisition module acquires the running state data during operation, to obtain cleaned running state data;

[0075] The inference module is used to input the cleaned running data into the AI ​​diagnostic model.

[0076] In one possible implementation, the device is specifically a cloud server; the device further includes: a training module.

[0077] The acquisition module is also used to acquire training sample data of multiple vehicles of the same type and training sample data of multiple vehicles of different types, and the motor driver is installed on the vehicle;

[0078] The training module is used to train a model based on a fault warning task using training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types, so as to obtain the preset AI diagnostic model.

[0079] In one possible implementation, the device is specifically a cloud server; the device further includes: a training module.

[0080] The acquisition module is also used to acquire training sample data of one vehicle or training sample data of multiple vehicles of the same type, and the motor driver is installed on the vehicle.

[0081] The training module is used to train a model based on a fault warning task using training sample data from one vehicle or training sample data from multiple vehicles of the same type, so as to obtain the preset AI diagnostic model.

[0082] In one possible implementation, the device is specifically a cloud server; the early warning module is used to generate fault early warning information based on the first anomaly parameter; and send the fault early warning information to the vehicle terminal device.

[0083] In one possible implementation, the device is specifically a vehicle terminal device;

[0084] When the cloud server sends the AI ​​diagnostic model to the vehicle terminal device, the acquisition module is also used to receive the AI ​​diagnostic model sent by the cloud server.

[0085] In one possible implementation, the device is specifically a cloud server; the device further includes: a training module.

[0086] The acquisition module is also used to acquire training sample data of multiple different types of vehicles, or training sample data of one vehicle, or training sample data of multiple vehicles of the same type.

[0087] The training module is used to extract data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations; and the pre-trained model is sent to the vehicle terminal device.

[0088] In one possible implementation, the device is specifically a vehicle terminal device; the device further includes: a training module.

[0089] The training module is used to receive a pre-trained model from a cloud server, wherein the pre-training module represents the data association relationship obtained by the cloud server after extracting training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type; and performs model parameter optimization processing on the pre-trained model according to preset label data to obtain the preset AI diagnostic model.

[0090] In one possible implementation, the runtime data includes: dynamic data collected by default during the operation of the motor driver; or,

[0091] The operational data includes dynamic data collected via the CAN bus of the controller area network.

[0092] In one possible implementation, the operational data includes dynamic data obtained by high-frequency sampling when the vehicle is in operation, and the motor driver is mounted on the vehicle.

[0093] In one possible implementation, the operational data includes at least one of the following: parameters of the motor driver, parameters of the cooling system, parameters of the motor, and parameters of the battery.

[0094] In one possible implementation, the parameters of the motor driver include at least one of the following: DC bus voltage, DC bus current, low-voltage power supply voltage, vehicle driving status command, vehicle status, motor driver temperature, rotor position electrical angle, effective value of motor driver three-phase current, motor driver three-phase current, sampled value of motor driver three-phase current, and motor driver derating status; and / or,

[0095] The parameters of the cooling system include at least one of the following: coolant flow rate, coolant temperature, oil pump target speed, oil pump status, oil pump actual speed, oil pump power supply voltage, oil pump power semiconductor device temperature, and oil pump current; and / or,

[0096] The parameters of the motor include at least one of the following: electromagnetic frequency, motor operating mode command, motor controller operating status, motor target torque command, motor current torque, motor target speed command, motor current speed, motor temperature, motor direct-axis voltage, motor direct-axis current setpoint, motor direct-axis current feedback value, motor quadrature-axis voltage, motor quadrature-axis current setpoint, motor quadrature-axis current feedback value, and motor current verification torque; and / or,

[0097] The parameters of the battery include at least one of the following: rated output voltage, battery capacity, and maximum output current.

[0098] In the second aspect of this application, the component modules of the fault warning device for the motor driver can also perform the steps described in the first aspect and various possible implementations, as detailed in the foregoing description of the first aspect and various possible implementations.

[0099] Thirdly, embodiments of this application also provide a fault early warning method for a motor driver, the method comprising:

[0100] Static data obtained by low-frequency sampling of the motor driver when the vehicle is stationary, wherein the motor driver is installed on the vehicle;

[0101] The static data is input into the AI ​​diagnostic model, and the AI ​​diagnostic model outputs the second anomaly parameter of the motor driver.

[0102] The motor driver is given a fault warning based on the second anomaly parameter.

[0103] In one possible implementation, the vehicle is in a stopped phase, including at least one of the following:

[0104] The vehicle is in the charging stage, the vehicle is in the starting stage, and the vehicle is in the off stage.

[0105] In one possible implementation, the step of providing fault warning for the motor driver based on the second anomaly parameter includes:

[0106] The AI ​​diagnostic model dynamically updates the preset abnormal conditions to obtain the updated abnormal conditions.

[0107] When the second anomaly parameter meets the updated anomaly condition, a fault warning is issued for the motor driver.

[0108] In one possible implementation, after acquiring static data obtained by low-frequency sampling of the motor driver when the vehicle is stationary, the method further includes:

[0109] The static data is cleaned to obtain cleaned static data;

[0110] The step of inputting the static data into a preset artificial intelligence (AI) diagnostic model includes:

[0111] The static data after the cleaning process is input into the AI ​​diagnostic model.

[0112] In one possible implementation, the method further includes:

[0113] The cloud server acquires training sample data from multiple vehicles of the same type and multiple vehicles of different types, and the motor driver is installed on the vehicle;

[0114] The cloud server uses training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types to train a model based on a fault warning task, so as to obtain the preset AI diagnostic model.

[0115] In one possible implementation, the method further includes:

[0116] The cloud server acquires training sample data for one vehicle or training sample data for multiple vehicles of the same type, and the motor driver is installed on the vehicle.

[0117] The cloud server uses training sample data from one vehicle or training sample data from multiple vehicles of the same type to train a model based on a fault warning task, so as to obtain the preset AI diagnostic model.

[0118] In one possible implementation, when the cloud server outputs a first anomaly parameter of the motor driver through the AI ​​diagnostic model, the step of providing a fault warning for the motor driver based on the second anomaly parameter includes:

[0119] The cloud server generates fault warning information based on the second anomaly parameter;

[0120] The cloud server sends the fault warning information to the vehicle terminal equipment.

[0121] In one possible implementation, when the cloud server sends the AI ​​diagnostic model to the vehicle terminal device, the method further includes: the vehicle terminal device receiving the AI ​​diagnostic model sent by the cloud server.

[0122] The step of inputting the static data into a preset artificial intelligence (AI) diagnostic model and outputting the second anomaly parameter of the motor driver through the AI ​​diagnostic model includes:

[0123] The vehicle terminal device inputs the static data into the AI ​​diagnostic model, and outputs the second anomaly parameter of the motor driver through the AI ​​diagnostic model.

[0124] In one possible implementation, the method further includes:

[0125] The cloud server acquires training sample data from multiple different types of vehicles, or training sample data from a single vehicle, or training sample data from multiple vehicles of the same type.

[0126] The cloud server extracts data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations.

[0127] The cloud server sends the pre-trained model to the vehicle terminal device.

[0128] In one possible implementation, the method further includes:

[0129] The vehicle terminal device receives a pre-trained model from the cloud server;

[0130] The vehicle terminal device performs model parameter optimization on the pre-trained model based on preset label data to obtain the preset AI diagnostic model.

[0131] Fourthly, embodiments of this application also provide a fault warning device for a motor driver, the device comprising:

[0132] An acquisition module is used to acquire static data obtained by low-frequency sampling of the motor driver when the vehicle is in a stopped state, wherein the motor driver is installed on the vehicle;

[0133] The inference module is used to input the static data into the AI ​​diagnostic model and output the second anomaly parameter of the motor driver through the AI ​​diagnostic model;

[0134] The early warning module is used to provide fault warnings to the motor driver based on the second anomaly parameter.

[0135] In one possible implementation, the vehicle is in a stopped phase, including at least one of the following:

[0136] The vehicle is in the charging stage, the vehicle is in the starting stage, and the vehicle is in the off stage.

[0137] In one possible implementation, the early warning module is used to dynamically update the preset abnormal conditions through the AI ​​diagnostic model to obtain the updated abnormal conditions; when the second abnormality parameter meets the updated abnormal conditions, a fault warning is issued to the motor driver.

[0138] In one possible implementation, the device further includes: a data processing module, used to clean the static data obtained by the acquisition module from low-frequency sampling of the motor driver when the vehicle is in a stopped state, to obtain cleaned static data.

[0139] The inference module is used to input the cleaned static data into the AI ​​diagnostic model.

[0140] In one possible implementation, the device is specifically a cloud server; the device further includes: a training module.

[0141] The acquisition module is also used to acquire training sample data of multiple vehicles of the same type and training sample data of multiple vehicles of different types, and the motor driver is installed on the vehicle;

[0142] The training module is used to train a model based on a fault warning task using training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types, so as to obtain the preset AI diagnostic model.

[0143] In one possible implementation, the device is specifically a cloud server; the device further includes: a training module.

[0144] The acquisition module is also used to acquire training sample data of one vehicle or training sample data of multiple vehicles of the same type, and the motor driver is installed on the vehicle.

[0145] The training module is used to train a model based on a fault warning task using training sample data from one vehicle or training sample data from multiple vehicles of the same type, so as to obtain the preset AI diagnostic model.

[0146] In one possible implementation, the device is specifically a cloud server; the early warning module is used to generate fault early warning information based on the second anomaly parameter; and send the fault early warning information to the vehicle terminal device.

[0147] In one possible implementation, the device is specifically a vehicle terminal device;

[0148] When the cloud server sends the AI ​​diagnostic model to the vehicle terminal device, the acquisition module is also used to receive the AI ​​diagnostic model sent by the cloud server.

[0149] In one possible implementation, the device is specifically a cloud server; the device further includes: a training module.

[0150] The acquisition module is also used to acquire training sample data of multiple different types of vehicles, or training sample data of one vehicle, or training sample data of multiple vehicles of the same type.

[0151] The training module is used to extract data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations; and the pre-trained model is sent to the vehicle terminal device.

[0152] In one possible implementation, the device is specifically a vehicle terminal device; the device further includes: a training module.

[0153] The training module is used to receive a pre-trained model from a cloud server, wherein the pre-training module represents the data association relationship obtained by the cloud server after extracting training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type; and performs model parameter optimization processing on the pre-trained model according to preset label data to obtain the preset AI diagnostic model.

[0154] Fifthly, embodiments of this application provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the methods described in the first or third aspect above.

[0155] Sixthly, embodiments of this application provide a computer program product containing instructions that, when run on a computer, cause the computer to perform the methods described in the first or third aspect above.

[0156] In a seventh aspect, embodiments of this application provide a communication device, which may include entities such as a cloud server, a vehicle terminal device, or a chip. The communication device includes: a processor and a memory; the memory is used to store instructions; the processor is used to execute the instructions in the memory, causing the communication device to perform the method as described in any one of the first or third aspects above.

[0157] Eighthly, this application provides a chip system including a processor for supporting a fault warning device in implementing the functions involved in the first or third aspect above, such as transmitting or processing data and / or information involved in the above methods. In one possible design, the chip system further includes a memory for storing necessary program instructions and data for the fault warning device. This chip system may be composed of chips or may include chips and other discrete devices.

[0158] Ninthly, this application provides a cloud server including a processor for supporting the fault warning device in implementing the functions involved in the first or third aspects described above, such as sending or processing data and / or information involved in the methods described above. In one possible design, the cloud server further includes a memory for storing necessary program instructions and data for the fault warning device.

[0159] In a tenth aspect, this application provides a vehicle terminal device, the chip system of which includes a processor for supporting a fault warning device in implementing the functions involved in the first or third aspect described above, such as sending or processing data and / or information involved in the above methods. In one possible design, the vehicle terminal device further includes a memory for storing necessary program instructions and data for the fault warning device.

[0160] Eleventhly, embodiments of this application also provide a fault warning system, which may include a cloud server as described in the ninth aspect and a vehicle terminal device as described in the tenth aspect.

[0161] In a twelfth aspect, embodiments of this application provide a vehicle-to-everything (V2X) device, which may include entities such as a V2X server, a roadside unit, a V2X communication device, or a chip. The V2X device includes a processor. Optionally, the V2X device further includes a memory; the memory is used to store instructions; the processor is used to execute the instructions in the memory, causing the V2X device to perform the method as described in any one of the first or third aspects above.

[0162] As can be seen from the above technical solutions, the embodiments of this application have the following advantages:

[0163] In this embodiment, the operating state data of the motor driver is first acquired, and then the operating state data is input into a preset AI diagnostic model. The AI ​​diagnostic model outputs a first anomaly parameter of the motor driver. The AI ​​diagnostic model is used to extract data features based on the operating state data and to perform decision-making inference based on these features. A fault warning is then issued for the motor driver based on the first anomaly parameter. Since the vehicle is in operation when the motor driver is running in this embodiment, a fault warning can be issued for the motor driver based on the operating state data. The operating state data of the motor driver is the data collected during the operation of the motor driver, and it reflects the true operating state of the motor driver. Fault warnings based on this operating state data can achieve early prediction of motor driver failure, thus realizing a fault warning for the motor driver. Attached Figure Description

[0164] Figure 1 A flowchart illustrating a fault warning method for a motor driver provided in an embodiment of this application;

[0165] Figure 2 A flowchart illustrating a fault warning method for a motor driver provided in an embodiment of this application;

[0166] Figure 3 A flowchart illustrating a fault warning method for a motor driver provided in an embodiment of this application;

[0167] Figure 4 A flowchart illustrating a fault warning method for a motor driver provided in an embodiment of this application;

[0168] Figure 5 A flowchart illustrating a fault warning method for a motor driver provided in an embodiment of this application;

[0169] Figure 6 A flowchart illustrating a fault warning method for a motor driver provided in an embodiment of this application;

[0170] Figure 7 This application provides a schematic diagram illustrating the composition of a cloud server and a vehicle terminal device according to an embodiment of the present application.

[0171] Figure 8 This is a schematic diagram of the execution flow of an AI diagnostic model provided in an embodiment of this application;

[0172] Figure 9 This is a schematic diagram of the execution flow of an AI diagnostic model provided in an embodiment of this application;

[0173] Figure 10This application provides a schematic diagram of the process for collecting runtime data and static data in an embodiment.

[0174] Figure 11a A schematic diagram of the composition structure of a fault early warning device for a motor driver provided in an embodiment of this application;

[0175] Figure 11b A schematic diagram of the composition structure of a fault early warning device for a motor driver provided in an embodiment of this application;

[0176] Figure 11c A schematic diagram of the composition structure of a fault early warning device for a motor driver provided in an embodiment of this application;

[0177] Figure 12 This is a schematic diagram of the composition structure of a fault warning device for a motor driver provided in an embodiment of this application. Detailed Implementation

[0178] This application provides a fault early warning method and apparatus for motor drivers, which are used to provide fault early warning for motor drivers.

[0179] The embodiments of this application will now be described with reference to the accompanying drawings.

[0180] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0181] This application provides a fault early warning method for a motor driver, which can provide fault early warning for the motor driver and improve the accuracy of fault early warning. The motor driver is installed in a vehicle, and is an electronic control device in the vehicle, such as a new energy vehicle, intelligent vehicle, or vehicle terminal equipment. The motor driver can also be called a motor control unit (MCU). The motor driver may include one or more power semiconductor devices, which are semiconductor devices that implement circuit switching functions. For example, a power semiconductor device can be a switching transistor. This application uses metal-oxide-semiconductor field-effect transistors (MOSFETs) as examples for illustration. It should be understood that the switching transistors can also be other semiconductor devices such as insulated-gate bipolar transistors (IGBTs), such as diodes. In subsequent embodiments, the switching transistors are mainly used in IGBTs for illustrative purposes. In this application embodiment, the motor driver controls the motor by controlling the motor's rotation angle and operating speed. For example, a motor driver can be driven by a relay or power transistor, or by a thyristor or a power MOSFET. There are various control requirements for a motor driver, such as controlling the motor's operating current and voltage, speed regulation, and forward and reverse rotation control of a DC motor.

[0182] In this embodiment of the application, the vehicle is in operation when the motor drive is running. Based on the operating status data of the motor drive, a fault warning can be given for the motor drive. Since the operating status data of the motor drive is the data collected when the motor drive is running, the operating status data can reflect the real operating status of the motor drive. Based on the operating status data, a fault warning can be given, which can realize the early prediction of motor drive failure and achieve fault warning for the motor drive.

[0183] The fault warning method for motor drives provided in this application embodiment is implemented by a fault warning device for motor drives (hereinafter referred to as the fault warning device). The fault warning device can store a pre-trained artificial intelligence (AI) diagnostic model. The AI ​​diagnostic model can be a diagnostic model obtained after model training using an AI diagnostic algorithm. The AI ​​diagnostic model has the ability to extract data features and performs decision reasoning based on the data features, outputting the anomaly parameter of the motor drive. By analyzing the anomaly parameter, it can be determined whether the motor drive has a fault, thereby providing a fault warning for the motor drive.

[0184] The fault warning device provided in this application can be implemented in various ways. The fault warning device can be a server independent of the vehicle or a terminal device integrated with the vehicle; this is not limited here. For example, the fault warning device can be a cloud server, which can also be called a cloud platform. The cloud server can store pre-trained AI diagnostic models, so that after the AI ​​diagnostic model completes decision reasoning, the cloud server can notify the vehicle terminal device of the reasoning results. Alternatively, the fault warning device can also be a vehicle terminal device, which can be the vehicle's controller or a terminal device integrated into the vehicle, also called an edge terminal device or edge intelligent device. The vehicle terminal device can obtain the reasoning results from the cloud server, thus enabling it to provide fault warnings for the motor drive. Another example is that the fault warning device can be a vehicle terminal device that can obtain a pre-trained model from the cloud server, and then generate an AI diagnostic model based on the pre-trained model, allowing the vehicle terminal device to use the AI ​​diagnostic model to provide fault warnings for the motor drive. This application does not limit the implementation method of the fault warning device.

[0185] Please refer to the following first. Figure 1 As shown in the figure, this application provides a fault early warning method for a motor driver, which mainly includes the following steps:

[0186] 101. Obtain the running state data of the motor driver during operation.

[0187] In this embodiment, the motor driver is installed in the vehicle. The motor driver needs to run during vehicle operation. The fault warning device can acquire the operating status data of the motor driver during operation. This operating status data can reflect the actual operating status of the motor driver. This embodiment does not limit the specific acquisition process of the operating status data of the motor driver or the acquisition range of the data items and data parameters included in the operating status data. For example, the operating status data can be comprehensively determined based on the type and function of the motor driver, the vehicle operating status, etc. The operating status data will be described in detail in subsequent embodiments.

[0188] For example, a fault warning device could be a cloud server. The vehicle terminal device could then collect operational status data of the motor drive during operation and transmit this data to the cloud server via a communication interface or wireless network. Alternatively, the fault warning device could be a vehicle terminal device equipped with a data measurement module that can collect operational status data of the motor drive.

[0189] In some embodiments of this application, the runtime data includes: dynamic data collected by default during motor driver operation; or,

[0190] The runtime data includes dynamic data collected via the controller area network (CAN) bus.

[0191] The operational data includes dynamic data collected by default during motor driver operation. This dynamic data is collected by default during motor driver operation and, in addition to its use in the vehicle's original control functions, can also be used for fault warning of the motor driver in this embodiment. Additionally, the operational data includes dynamic data collected via the CAN bus. Since the motor controller is connected to the CAN bus, dynamic data can be collected via the CAN bus to obtain the aforementioned operational data. This embodiment is not limited to the two data acquisition methods described above. These methods do not require hardware upgrades to the vehicle or motor driver boards to obtain the operational data of the motor driver, thus improving the efficiency of operational data acquisition. This embodiment does not limit the specific acquisition process of the motor driver's operational data or the range of data items and parameters included in the operational data.

[0192] In some embodiments of this application, the operational data includes: dynamic data obtained by high-frequency sampling when the vehicle is in operation, and the motor driver is installed on the vehicle.

[0193] The vehicle has multiple states, such as being in an operating phase, which refers to the stage after the vehicle is started, or being in motion. The dynamic data obtained by high-frequency sampling when the vehicle is in the operating phase can be the aforementioned operating state data. Subsequent embodiments will provide examples of the high-frequency sampling data items and data volume. High-frequency sampling refers to data acquisition at a frequency greater than a preset frequency threshold; however, the specific value of the frequency threshold is not limited in this embodiment.

[0194] In some embodiments of this application, the operational data includes at least one of the following: parameters of the motor driver, parameters of the cooling system, parameters of the motor, and parameters of the battery.

[0195] In this application embodiment, the running state data of the motor driver can be implemented in various ways. For example, data can be collected from the motor driver to obtain its parameters, or data can be collected from the cooling system to obtain its parameters, or data can be collected from the motor itself to obtain its parameters, or data can be collected from the battery to obtain its parameters. This application embodiment does not limit the specific parameter names and contents of the parameters of the motor driver, cooling system, motor, and battery.

[0196] It should be noted that, in this embodiment, the parameters of the motor driver, cooling system, motor, and battery can be obtained through data acquisition. Since the parameters of the motor driver, cooling system, motor, and battery are all data collected during the operation of the motor driver, any one of these parameters can reflect the actual operating status of the motor driver. Based on these parameters, fault warning can be performed to predict motor driver failure in advance, thus achieving fault warning for the motor driver.

[0197] Furthermore, in some embodiments of this application, the parameters of the motor driver include at least one of the following: DC bus voltage, DC bus current, low-voltage power supply voltage, vehicle driving status command, vehicle status, motor driver temperature, rotor position electrical angle, effective value of motor driver three-phase current, motor driver three-phase current, sampled value of motor driver three-phase current, and motor driver derating status; and / or,

[0198] The parameters of the cooling system include at least one of the following: coolant flow rate, coolant temperature, oil pump target speed, oil pump status, oil pump actual speed, oil pump supply voltage, oil pump power semiconductor device temperature, and oil pump current; and / or,

[0199] The motor parameters include at least one of the following: electromagnetic frequency, motor operating mode command, motor controller operating status, motor target torque command, motor current torque, motor target speed command, motor current speed, motor temperature, motor direct-axis voltage, motor direct-axis current setpoint, motor direct-axis current feedback value, motor quadrature-axis voltage, motor quadrature-axis current setpoint, motor quadrature-axis current feedback value, and motor current verification torque; and / or,

[0200] The battery parameters include at least one of the following: rated output voltage, battery capacity, and maximum output current.

[0201] Among the parameters of the motor driver, the three-phase current of the motor driver can be the current of phases U, V and W.

[0202] It should be noted that detailed descriptions of the parameters of the motor driver, cooling system, motor, and battery are provided in subsequent embodiments.

[0203] 102. Input the running state data into the preset artificial intelligence (AI) diagnostic model, and output the first anomaly parameter of the motor driver through the AI ​​diagnostic model. The AI ​​diagnostic model is used to extract data features based on the running state data and to make decision-making inferences based on the data features.

[0204] In this embodiment, the fault diagnosis device can pre-train an AI diagnostic model. After acquiring operational data, the fault diagnosis device uses the AI ​​diagnostic model for decision-making and reasoning. Specifically, operational data is input into the AI ​​diagnostic model, which extracts data features based on the operational data. These data features are obtained in advance by the AI ​​diagnostic model based on the input operational data and serve as the basis for the AI ​​diagnostic model's decision-making and reasoning. The AI ​​diagnostic model can be a machine learning model, such as a pre-trained neural network model or other machine learning models, such as linear regression models or decision tree models. The AI ​​diagnostic model performs decision-making and reasoning based on the extracted data features. In this embodiment, the data content included in the data features is not limited. After the AI ​​diagnostic model performs decision-making and reasoning, it can output a first anomaly parameter of the motor drive. This first anomaly parameter is defined to distinguish it from other anomaly parameters (such as a second anomaly parameter) that appear in subsequent embodiments. The first anomaly parameter is used to measure the abnormal condition of the motor drive and can also be called a first anomaly factor score, a first anomaly rating, etc.

[0205] It should be noted that the fault diagnosis device performs the aforementioned step 102. Specifically, the fault diagnosis device can be a cloud server or a vehicle terminal device.

[0206] In some embodiments of this application, after obtaining the running state data of the motor driver in step 101, the method provided in this application embodiment may further include the following steps:

[0207] The runtime data is cleaned to obtain cleaned runtime data.

[0208] After acquiring multiple operational data points, the fault diagnosis device can first clean the operational data, for example, by removing data that meets the cleaning criteria. These criteria could include incomplete operational data or a certain indicator exceeding a specified limit.

[0209] After cleaning the runtime data, step 102 above, which involves inputting the runtime data into a preset AI diagnostic model, includes:

[0210] The cleaned operational data is input into the AI ​​diagnostic model.

[0211] The fault diagnosis device acquires the cleaned operational data and then inputs it into the AI ​​diagnostic model. The AI ​​diagnostic model extracts data features from the cleaned operational data. By cleaning the operational data before inputting it into the AI ​​diagnostic model, the inference efficiency of the AI ​​diagnostic model can be improved.

[0212] 103. Provide fault warnings for the motor driver based on the first anomaly parameter.

[0213] In this embodiment, after the AI ​​diagnostic model outputs a first anomaly parameter, the operating status of the motor driver can be measured based on this parameter to identify whether the motor driver is experiencing any abnormalities. When an abnormality occurs, it is determined that the motor driver is faulty, and a fault warning is issued to the motor driver, thereby achieving early prediction of motor driver failure. Here, fault warning refers to predicting that a fault will occur in the future a certain period of time before it actually happens and issuing an alarm signal. The method provided in this embodiment can be applied to automotive electric drive systems, diagnosing failures caused by the motor driver in advance and issuing warnings, allowing users to perform maintenance before a fault occurs and avoiding vehicle breakdowns during driving.

[0214] In some embodiments of this application, step 103 provides a fault warning for the motor driver based on a first anomaly parameter, including:

[0215] The AI ​​diagnostic model dynamically updates the preset abnormal conditions to obtain the updated abnormal conditions.

[0216] When the first anomaly parameter meets the updated anomaly condition, a fault warning is issued for the motor driver.

[0217] The fault diagnosis device can preset abnormal conditions, which can be dynamically updated through an AI diagnostic model to ensure that the abnormal conditions can be used for fault diagnosis and to guarantee the accuracy of the fault diagnosis. After dynamically updating the abnormal conditions, it can be determined whether the first abnormality degree meets the dynamically updated abnormal conditions. When the first abnormality degree parameter meets the updated abnormal conditions, it indicates that the motor driver may have a fault, and a fault warning is issued to the motor driver. In this embodiment, by dynamically updating the abnormal conditions, the accuracy of the fault warning can be further improved.

[0218] As illustrated by the foregoing embodiments, the process begins by acquiring operational status data of the motor drive during operation. This operational status data is then input into a preset AI diagnostic model, which outputs a first anomaly parameter for the motor drive. The AI ​​diagnostic model extracts data features from the operational status data and performs decision-making inference based on these features. A fault warning is then issued for the motor drive based on the first anomaly parameter. Since the vehicle is in operation when the motor drive is running in this embodiment, a fault warning can be issued for the motor drive based on its operational status data. This operational status data, collected during operation, reflects the actual operating state of the motor drive. By issuing a fault warning based on this operational status data, early prediction of motor drive failure can be achieved, thus enabling early warning of motor drive failures.

[0219] The foregoing embodiments of this application illustrate fault warning based on the operating state data of the motor driver. The following describes a fault warning scheme based on static data when the vehicle is stationary, provided by embodiments of this application. Please refer to [link to relevant documentation]. Figure 2 As shown in the figure, this application provides a fault early warning method for a motor driver, which mainly includes the following steps:

[0220] 201. Obtain static data obtained by low-frequency sampling of the motor driver when the vehicle is stationary. The motor driver is installed on the vehicle.

[0221] The vehicle has multiple states, such as when it is in a stopped state, which can also be called a stationary state.

[0222] When the vehicle is stationary, low-frequency sampling is performed on the motor driver to obtain static data. Subsequent embodiments will illustrate the data items and data volume of the low-frequency sampling. Low-frequency sampling refers to data acquisition at a frequency lower than a preset frequency threshold; however, the value of this frequency threshold is not limited in this embodiment.

[0223] For example, the fault diagnosis device could be a vehicle terminal device. When the vehicle is stationary, the vehicle terminal device can perform low-frequency sampling of the motor driver to obtain static data, which can then be used for decision-making and inference in an AI diagnostic model. Alternatively, the fault diagnosis device could be a cloud server. When the vehicle is stationary, the vehicle terminal device can perform low-frequency sampling of the motor driver to obtain static data, which can then be sent to the cloud server, allowing the cloud server to receive the static data from the vehicle terminal device.

[0224] In some embodiments of this application, the vehicle is in a stopped phase, including at least one of the following: the vehicle is in a charging phase, the vehicle is in an ignition phase, and the vehicle is in a shutdown phase.

[0225] In particular, static data is sampled during the vehicle charging phase, vehicle ignition phase, and vehicle shutdown phase, so that the fault diagnosis device can provide emergency response functions in the event of a network outage based on the static data.

[0226] 202. Input static data into the AI ​​diagnostic model, and output the second anomaly parameter of the motor driver through the AI ​​diagnostic model.

[0227] In this embodiment, the fault diagnosis device can pre-train an AI diagnostic model. After acquiring static data, the fault diagnosis device uses the AI ​​diagnostic model for decision-making and reasoning. Specifically, static data is input into the AI ​​diagnostic model, which extracts data features based on the static data. These data features are obtained in advance by the AI ​​diagnostic model based on the input static data and serve as the basis for the AI ​​diagnostic model's decision-making and reasoning. The AI ​​diagnostic model can be a machine learning model, such as a pre-trained neural network model, or other machine learning models, such as linear regression models or decision tree models. The AI ​​diagnostic model performs decision-making and reasoning based on the extracted data features. In this embodiment, the data content included in the data features is not limited. After the AI ​​diagnostic model performs decision-making and reasoning, it can output a second anomaly parameter for the motor drive. The second anomaly parameter is used to measure the abnormal condition of the motor drive and can also be called a second anomaly factor score, second anomaly rating, etc.

[0228] 203. Provide fault warnings for the motor driver based on the second anomaly parameter.

[0229] In this embodiment, after the AI ​​diagnostic model outputs a second anomaly parameter, the operating status of the motor driver can be measured based on this parameter to identify whether the motor driver is experiencing any abnormalities. When an abnormality occurs, it is determined that the motor driver is faulty, and a fault warning is issued to the motor driver, thereby achieving early prediction of motor driver failure. The method provided in this embodiment can be applied to automotive electric drive systems, diagnosing and issuing warnings for failures caused by the motor driver in advance, allowing users to perform maintenance before a fault occurs and avoiding situations such as vehicle breakdowns during driving.

[0230] It should be noted that steps 201 to 203 provided in the embodiments of this application can be independent of... Figure 1 In the embodiment shown, steps 201 to 203 can also be performed in... Figure 1 Steps 101 to 103 shown can be executed afterward, or steps 201 to 203 can also be performed afterward. Figure 1 The steps 101 to 103 shown are executed before this step, and are not limited here. For example, the fault diagnosis device can acquire the aforementioned static data and operational data, input both static data and operational data into the AI ​​diagnosis model, extract data features based on the static data and operational data, perform decision reasoning based on the data features, output a third anomaly parameter, and finally provide a fault warning for the motor driver based on the third anomaly parameter.

[0231] As can be seen from the examples in the foregoing embodiments, in this application embodiment, static data of the motor driver is acquired when the vehicle is stationary. Based on this static data, a fault warning can be given for the motor driver. The static data of the motor driver is the data collected from the motor driver when the vehicle is stationary. The static data can reflect the true stationary state of the motor driver. Based on this static data, a fault warning can be given, which can realize the early prediction of motor driver failure and achieve fault warning for the motor driver.

[0232] The foregoing Figure 1 and Figure 2 The method for fault warning of a motor drive is illustrated by an example of a fault diagnosis device. The following describes an interaction process between a cloud server and a vehicle terminal device in an embodiment of this application. Figure 3 As shown in the figure, this application provides a fault early warning method for a motor driver, which mainly includes the following steps:

[0233] 301. The cloud server obtains training sample data from multiple vehicles of the same type and multiple vehicles of different types, and the motor driver is installed on the vehicle.

[0234] The cloud server can interact with multiple vehicle terminal devices, and each vehicle can be equipped with one or more motor drives. The cloud server can acquire multiple training sample data sets, which can be used for model training. For example, these training sample data sets could be from multiple vehicles of the same type or from multiple vehicles of different types, where "type" refers to the vehicle's type. There are no restrictions on the data type or content of the training sample data.

[0235] 302. The cloud server uses training sample data from multiple vehicles of the same type and multiple vehicles of different types to train a model based on the fault warning task, so as to obtain a preset AI diagnostic model.

[0236] In this embodiment, the cloud server acquires training sample data from multiple vehicles of the same type and multiple vehicles of different types. The cloud server can predefine the training task as a fault warning task. The cloud server uses the training sample data from multiple vehicles of the same type and multiple vehicles of different types to train the model. The specific training process of the model is not described in detail in this embodiment. After the model training is completed, a preset AI diagnostic model can be obtained. For example, the cloud server first uses the training sample data from multiple vehicles of the same type to train the model, and then generalizes it to the training sample data from different types of vehicles. Finally, the model training can be completed, and the preset AI diagnostic model can be output.

[0237] 303. The cloud server obtains the running status data of the motor driver sent by the vehicle terminal device.

[0238] 304. The cloud server inputs the running state data into the preset artificial intelligence (AI) diagnostic model, and outputs the first anomaly parameter of the motor driver through the AI ​​diagnostic model. The AI ​​diagnostic model is used to extract data features based on the running state data and to make decision-making inferences based on the data features.

[0239] In this embodiment, the cloud server can use an AI diagnostic model to perform decision-making reasoning and output the first anomaly parameter of the motor driver through the AI ​​diagnostic model. The implementation methods of steps 303 to 304 are similar to those of steps 101 to 102 in the aforementioned embodiments, and are not limited here.

[0240] 305. The cloud server generates fault warning information based on the first anomaly parameter.

[0241] 306. The cloud server sends fault warning information to the vehicle terminal equipment.

[0242] In this process, after the cloud server outputs the first anomaly parameter through the AI ​​diagnostic model, the cloud server generates fault warning information based on the first anomaly parameter. This fault warning information can be the result of the cloud server's reasoning and decision-making based on the first anomaly parameter. The cloud server sends the fault warning information to the vehicle terminal device so that the vehicle terminal device can provide a fault warning for the motor driver based on the fault warning information.

[0243] 307. The vehicle terminal equipment receives fault warning information from the cloud server.

[0244] 308. The vehicle terminal equipment provides fault warnings to the motor driver based on the fault warning information.

[0245] It should be noted that in this embodiment, the cloud server uses multiple training sample data, which may include training sample data of multiple vehicles of the same type and training sample data of multiple vehicles of different types. The resulting AI diagnostic model is a global AI model, which is designed for specific fault warning tasks. The model is trained based on all collected vehicle terminal device side data. Based on the model, the end-side inference results of each vehicle terminal device are given as a decision strategy and sent to the vehicle terminal devices. The vehicle terminal devices do not perform training inference but directly execute the inference results of the cloud server. Therefore, fault warnings for motor drives can be completed through the interaction between the cloud server and the vehicle terminal devices.

[0246] As can be seen from the examples in the foregoing embodiments, the cloud server in this application embodiment has model training capability and decision reasoning capability. The vehicle terminal device can perform fault warning according to the decision results of the cloud server. Through the interaction between the cloud server and the vehicle terminal device, fault warning for the motor drive can be realized.

[0247] like Figure 4 As shown in the figure, this application provides a fault early warning method for a motor driver, which mainly includes the following steps:

[0248] 401. The cloud server obtains training sample data from multiple vehicles of the same type and multiple vehicles of different types, and the motor driver is installed on the vehicle.

[0249] 402. The cloud server uses training sample data from multiple vehicles of the same type and multiple vehicles of different types to train a model based on the fault warning task, so as to obtain a preset AI diagnostic model.

[0250] The implementation methods of steps 401 to 402 are similar to those of steps 301 to 302 in the foregoing embodiments. For details, please refer to the foregoing embodiment descriptions. No limitations are made here.

[0251] 403. The cloud server sends the AI ​​diagnostic model to the vehicle terminal equipment.

[0252] In this embodiment of the application, after the cloud server completes the training of the AI ​​diagnostic model, the cloud server can send the trained AI diagnostic model to the vehicle terminal device.

[0253] 404. The vehicle terminal equipment receives the AI ​​diagnostic model sent by the cloud server.

[0254] In this embodiment, the vehicle terminal device does not have model training capabilities. The vehicle terminal device can receive an AI diagnostic model from a cloud server, and thus the vehicle terminal device can use the AI ​​diagnostic model to make decision-making inferences. Specifically, the vehicle terminal device can execute subsequent steps 405 to 407.

[0255] 405. The vehicle terminal equipment acquires the running status data of the motor driver during operation.

[0256] 406. The vehicle terminal equipment inputs the operating status data into the preset AI diagnostic model, and outputs the first anomaly parameter of the motor driver through the AI ​​diagnostic model. The AI ​​diagnostic model is used to extract data features based on the operating status data and to make decision-making inferences based on the data features.

[0257] 407. The vehicle terminal equipment provides a fault warning for the motor driver based on the first anomaly parameter.

[0258] The implementation methods of steps 405 to 407 are similar to those of steps 101 to 103 in the aforementioned embodiments, as detailed in the aforementioned embodiment descriptions, and are not limited here.

[0259] It should be noted that in this embodiment, the cloud server uses multiple training sample data, which may include training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types. The resulting AI diagnostic model is a global AI model, which is designed for specific fault warning tasks. This model is trained based on all collected vehicle terminal device data. The cloud server sends the trained AI diagnostic model to the vehicle terminal device. The vehicle terminal device uses this model to provide its own edge-side inference results. The vehicle terminal device does not perform training but directly uses the AI ​​diagnostic model for inference and executes the inference results. Therefore, fault warnings for the motor drive can be completed through the interaction between the cloud server and the vehicle terminal device.

[0260] As can be seen from the examples in the foregoing embodiments, the cloud server in this application embodiment has model training capability, the vehicle terminal device has decision reasoning capability, the vehicle terminal device can receive the AI ​​diagnostic model from the cloud server, and thus the vehicle terminal device can output decision results according to the AI ​​diagnostic model to provide fault warning for the motor drive. Through the interaction between the cloud server and the vehicle terminal device, fault warning for the motor drive can be realized.

[0261] like Figure 5 As shown in the figure, this application provides a fault early warning method for a motor driver, which mainly includes the following steps:

[0262] 501. The cloud server obtains training sample data for one vehicle or multiple vehicles of the same type, and the motor driver is installed on the vehicle.

[0263] The cloud server can interact with one vehicle terminal device or multiple vehicle terminal devices of the same type, and each vehicle can be equipped with one or more motor drives. The cloud server can acquire multiple training sample data, which can be used for model training. For example, the multiple training sample data can be the training sample data of one vehicle or the training sample data of multiple vehicles of the same type, where "type" refers to the vehicle type. There are no restrictions on the data type and content of the training sample data.

[0264] 502. The cloud server uses training sample data from one vehicle or training sample data from multiple vehicles of the same type to train a model based on the fault warning task, so as to obtain a preset AI diagnostic model.

[0265] In this embodiment, the cloud server acquires training sample data from one vehicle or multiple vehicles of the same type. The cloud server can predefine the training task as a fault warning task. The cloud server uses the training sample data from one vehicle or multiple vehicles of the same type to train the model. The specific training process of the model is not described in detail in this embodiment. After the model training is completed, a preset AI diagnostic model can be obtained. For example, the cloud server first uses the training sample data from one vehicle to train the model, and then generalizes it using the training sample data from multiple vehicles of the same type. Finally, the model training can be completed, and the preset AI diagnostic model can be output.

[0266] 503. The cloud server obtains the running status data of the motor driver sent by the vehicle terminal device.

[0267] 504. The cloud server inputs the running state data into the preset artificial intelligence (AI) diagnostic model, and outputs the first anomaly parameter of the motor driver through the AI ​​diagnostic model. The AI ​​diagnostic model is used to extract data features based on the running state data and to make decision-making inferences based on the data features.

[0268] 505. The cloud server generates fault warning information based on the first anomaly parameter.

[0269] 506. The cloud server sends fault warning information to the vehicle terminal equipment.

[0270] 507. The vehicle terminal equipment receives fault warning information from the cloud server.

[0271] 508. The vehicle terminal equipment provides fault warnings to the motor driver based on the fault warning information.

[0272] The implementation methods of steps 503 to 508 are similar to those of steps 303 to 308 in the aforementioned embodiments, as detailed in the aforementioned embodiment descriptions, and are not limited here.

[0273] It should be noted that in this embodiment, the cloud server uses multiple training sample data, which may include training sample data of one vehicle or training sample data of multiple vehicles of the same type. The resulting AI diagnostic model is a local AI model, which is designed for a specific fault warning task. This local AI model may be a certain model of vehicle terminal device or even a specific vehicle terminal device. Based on this model, each edge-side inference result is given and sent to the vehicle terminal device as a decision strategy. The vehicle terminal device does not perform training inference but directly executes the inference result of the cloud server. Therefore, fault warning for the motor drive can be completed through the interaction between the cloud server and the vehicle terminal device.

[0274] As can be seen from the examples in the foregoing embodiments, the cloud server in this application embodiment has model training capability and decision reasoning capability. The vehicle terminal device can perform fault warning according to the decision results of the cloud server. Through the interaction between the cloud server and the vehicle terminal device, fault warning for the motor drive can be realized.

[0275] like Figure 6 As shown in the figure, this application provides a fault early warning method for a motor driver, which mainly includes the following steps:

[0276] 601. The cloud server obtains training sample data of multiple different types of vehicles, or training sample data of one vehicle, or training sample data of multiple vehicles of the same type.

[0277] The cloud server can interact with different types of vehicle terminal devices, or a single vehicle terminal device, or multiple vehicle terminal devices of the same type. Each vehicle can be equipped with one or more motor drivers. The cloud server can acquire multiple training sample data sets, which can be used to extract data associations. For example, the multiple training sample data sets can be training sample data from multiple different types of vehicles, or training sample data from a single vehicle, or training sample data from multiple vehicles of the same type, where "type" refers to the vehicle type. There are no restrictions on the data type or content of the training sample data.

[0278] 602. The cloud server extracts data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations.

[0279] In this process, the cloud server is not task- or function-specific. It extracts data associations from training sample data of multiple different types of vehicles, or from training sample data of a single vehicle, or from training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The data associations can be the relationships between various data dimensions and their features in multiple training sample data within a certain time sequence and operating condition.

[0280] 603. The cloud server sends the pre-trained model to the vehicle terminal equipment.

[0281] In this embodiment of the application, after obtaining the pre-trained model, the cloud server can send the pre-trained model to the vehicle terminal device.

[0282] 604. The vehicle terminal equipment receives the pre-trained model from the cloud server.

[0283] 605. The vehicle terminal equipment performs model parameter optimization on the pre-trained model based on the preset tag data to obtain the preset AI diagnostic model.

[0284] In this embodiment, the cloud server does not need to train the AI ​​diagnostic model. Instead, it only needs to extract data relationships to obtain a pre-trained model, which it then sends to the vehicle terminal device. The vehicle terminal device optimizes the model parameters based on preset label data; this optimization can also be called fine-tuning. In other words, the vehicle terminal device can train the model to obtain the preset AI diagnostic model. For example, after receiving the pre-trained model, the vehicle terminal device fine-tunes the model based on the fault warning task and the preset label data, achieving the fault warning task with low resource consumption.

[0285] 606. The vehicle terminal equipment acquires the running status data of the motor driver during operation.

[0286] 607. The vehicle terminal equipment inputs the operating status data into the preset AI diagnostic model, and outputs the first anomaly parameter of the motor driver through the AI ​​diagnostic model. The AI ​​diagnostic model is used to extract data features based on the operating status data and to make decision-making inferences based on the data features.

[0287] 608. The vehicle terminal equipment provides a fault warning for the motor driver based on the first anomaly parameter.

[0288] The implementation methods of steps 606 to 608 are similar to those of steps 101 to 103 in the aforementioned embodiments, as detailed in the aforementioned embodiment descriptions, and are not limited here.

[0289] It should be noted that in this embodiment, the cloud server uses multiple training sample data. The cloud server can obtain a pre-trained model and distribute the pre-trained model to the vehicle terminal device as the initialization parameters for training. The vehicle terminal device optimizes the model parameters based on the small amount of collected data, i.e., fine-tuning or transfer learning process, thereby obtaining the AI ​​diagnostic model. The vehicle terminal device provides the edge-side inference result for each vehicle terminal device based on the model. The vehicle terminal device directly uses the AI ​​diagnostic model to perform inference and executes the inference result. Therefore, through the interaction between the cloud server and the vehicle terminal device, fault warning for the motor drive can be completed.

[0290] As illustrated by the examples in the foregoing embodiments, in this application embodiment, the cloud server sends a pre-trained model to the vehicle terminal device, and the vehicle terminal device generates an AI diagnostic model based on the pre-trained model. The vehicle terminal device has decision-making and reasoning capabilities, so the vehicle terminal device can output decision results based on the AI ​​diagnostic model and provide fault warnings for the motor drive. Through the interaction between the cloud server and the vehicle terminal device, fault warnings for the motor drive can be achieved.

[0291] To facilitate a better understanding and implementation of the above-described solutions in the embodiments of this application, specific examples of corresponding application scenarios are provided below.

[0292] This application proposes a method for early warning of faults in motor drives. This method is applicable not only to motor drives but also to all power electronic components using power semiconductor modules, such as chargers and DC / DC converters. The technical solution provided in this application enables intelligent diagnosis of the entire motor drive. The following explanation uses a power semiconductor device as a specific example of a motor drive. Operating data of the power semiconductor device is collected during vehicle operation. An AI diagnostic model is used to determine data characteristics and score the anomalies. Power semiconductor devices with excessive anomalies are identified, thus enabling early prediction of failure of individual power semiconductor devices.

[0293] Please see as follows Figure 7 The diagram shown is a schematic representation of the structural composition of a cloud server and a vehicle terminal device according to an embodiment of this application. The cloud server, also referred to as the cloud, includes at least one of the following modules: a data processing module, a data storage module, a model training module, and a cloud inference module. The vehicle terminal device, also referred to as the edge-side or board-side, includes at least one of the following modules: a data measurement module, a fine-tuning module, a strategy execution module, and an edge-side inference module.

[0294] The data measurement module is used to collect training sample data of power semiconductor devices.

[0295] The data storage module is used to store the massive amounts of data reported by the storage side, facilitating subsequent processing.

[0296] The data processing module is used to clean and process massive amounts of data to facilitate the training of subsequent AI diagnostic models.

[0297] The model training module is used for fault warning or state calculation. The following will use a fault warning scenario as an example to illustrate the process.

[0298] The model training module on the cloud server has the following three implementation methods:

[0299] 1. The AI ​​diagnostic model is a general-purpose model. All vehicles share the same model, which is designed for specific tasks or functions, such as fault warning or capacity estimation. The model can be directly used for result inference in these scenarios. Feature information (such as fault or capacity-related features) is extracted from training sample vehicles to train a model. The cloud inference module can then use this model to perform inference and prediction on the data of each vehicle on the cloud server to obtain the inference result. The inference result (such as fault warning) is used as a policy and sent to the vehicle terminal device. The policy execution module in the vehicle terminal device controls and processes the data according to the policy of the cloud server.

[0300] 2. The AI ​​diagnostic model is a local model. A single model may be used for a class of vehicles, or a separate model may be used for a single vehicle. Similar to the aforementioned general models, these local models are task-specific. Features are extracted from data on a specific class of vehicles or each vehicle on a cloud server, and the AI ​​diagnostic model is trained. Based on these models, the cloud server performs inference and prediction on the corresponding vehicles to obtain inference results. These inference results (e.g., fault warnings) are used as policies and sent to the vehicle's terminal device. The policy execution module in the vehicle's terminal device controls and processes the data according to the policies from the cloud server.

[0301] 3. Obtaining AI diagnostic models through pre-trained models. Pre-trained models are not specific to any particular task or function. They learn the relationships between various data dimensions and their features from massive amounts of data within a given time series and operating condition. The fine-tuning module of the vehicle terminal equipment can then fine-tune the pre-trained model using labeled data to obtain an AI diagnostic model. This model enables tasks such as fault warning and capacity estimation with low resource consumption. Taking the automotive field as an example, cloud servers store vast amounts of data on time series, operating conditions, driving behavior, vehicle location, total current, total voltage, temperature, motor status, alarm types and levels, and fault states. The pre-trained model learns the relationships between these dimensions and their derived features through self-supervised learning. When performing tasks such as fault warning, the vehicle terminal equipment only needs to fine-tune the pre-trained model using its own labeled alarm data to obtain an AI diagnostic model for the vehicle's fault warning task. This model contains both general information learned from big data and its own unique information. The general information refers to the data distribution of various measurements learned by the cloud server from multiple operating vehicles, reflecting the interrelationships between these measurements under big data conditions. Unique information refers to the measurement values ​​of the vehicle terminal equipment. After fine-tuning or transfer learning by combining the general information of the pre-trained model, a unique AI diagnostic model can be obtained on the edge.

[0302] The cloud-based inference module executes the inference function of the cloud server, providing inference results for tasks such as fault warning or capacity estimation based on a general or local model trained in the cloud.

[0303] The fine-tuning module performs edge-side fine-tuning (i.e., transfer learning) functions. Based on the pre-trained model in the cloud, it combines data from the vehicle terminal device to fine-tune the model, resulting in an AI diagnostic model for a specific task (such as a fault warning task).

[0304] The edge-side inference module performs edge-side inference functions, supporting inference in two scenarios: One is where a lightweighted general or local model from the cloud server is deployed to the vehicle terminal device, which then performs inference based on the AI ​​diagnostic model sent from the cloud server. The other is where the vehicle terminal device performs inference after obtaining an AI diagnostic model based on a cloud-based pre-trained model and fine-tuning (transfer learning).

[0305] The strategy execution module is used to process the inference results of security warnings (such as fault warning tasks) or state estimations (such as capacity estimation tasks).

[0306] As can be seen from the above examples, the embodiments of this application may adopt cloud-based training and inference and edge-based training and inference, or only cloud-based inference, or only edge-based inference, without limiting the specific implementation method.

[0307] This embodiment uses the default operating state data of the motor driver as the input of the AI ​​diagnostic model to calculate the abnormal factor score of the motor driver. The abnormal factor score is used to determine the risk level of the motor driver's future failure. For products with high risk levels, a warning signal is issued before the actual failure occurs, so that users can detect and repair in advance and avoid damage during vehicle operation.

[0308] like Figure 8 The diagram shown illustrates the execution flow of an AI diagnostic model provided in an embodiment of this application. For example, the input data for the AI ​​diagnostic model includes: bus voltage V_dc, three-phase current I_phase, IGBT temperature T_jav, coolant flow rate L, coolant temperature T_fluid, and electromagnetic frequency f_em.

[0309] Specifically, the input runtime data is shown in Table 1 below. This data is the default sampled data during MCU runtime and can be directly obtained from the CAN bus. The specifications of the collected runtime data may differ for different vehicles. This is just an example and is not a limitation. Any data that can be collected from the CAN bus can be used as input data for the AI ​​diagnostic model.

[0310] Table 1 shows some examples of the collected runtime data. More test items can be added, corresponding to the failure modes that need to be detected.

[0311] Data collection projects DC bus voltage DC bus current Drive motor operating mode command Current torque of drive motor Current speed of drive motor Drive motor temperature Ud (Voltage controlling the d-axis of the motor) Uq (Voltage controlling the q-axis of the motor) MCU IGBT temperature (U phase) MCU IGBT temperature (V phase) MCU IGBT Temperature (W Phase) U-phase current sampling value V-phase current sampling value W-phase current sampling value

[0312] The AI ​​diagnostic model performs decision-making inference based on the above input data and outputs the first anomaly parameter. For details on the decision-making inference process of the AI ​​diagnostic model, please refer to the preceding text. Figure 7 The explanation of the model reasoning process shown will not be repeated here.

[0313] In other embodiments of this application, additional sampling sensors or circuits are used in the motor driver to extract specific key data as input data for the AI ​​diagnostic model, calculate the anomaly factor score of the power semiconductor device, and use the anomaly factor score to determine the risk level of future failure of the power semiconductor device. For products with high risk levels, a warning signal is issued before the actual failure occurs, so that users can detect and repair in advance and avoid damage during vehicle operation.

[0314] like Figure 9 The diagram illustrates an execution flow of an AI diagnostic model. The input data for this model may include: breakdown voltage Brv_ce, CE leakage current I_ces, GE leakage current I_ges, threshold voltage Vth, IGBT saturation voltage drop V_cesat, diode forward voltage drop V_f, IGBT parasitic capacitances Cies, Coes, Cres, and module switching losses Eon, Eoff, Err. The specific meanings of these input data will not be described in detail in this embodiment.

[0315] The test circuit used in the embodiments of this application can be implemented in various ways. For example, the test circuit can be an Iges test circuit, a Vth test circuit, a Bvces and Ices test circuit, an Ices low-end test circuit, a Vcesat and Vf test circuit. The type of test circuit is not limited in the embodiments of this application.

[0316] The input data for the AI ​​diagnostic model is shown in Tables 2 and 3 below for IGBT and MOS power semiconductor devices, respectively. These data are not the default data sampled by the MCU during runtime. Additional sensors or sampling circuits can be added to the MCU to achieve data acquisition.

[0317] Table 2 lists some of the test items for the IGBT module. Additional test items can be added to the actual test list, and each test item corresponds to a failure mode that needs to be detected.

[0318]

[0319] Table 3 lists some of the test items for MOSFETs. Additional items can be added in actual testing. The test items correspond to the failure modes that need to be detected.

[0320]

[0321] like Figure 10 The diagram illustrates a process for acquiring operational and static data according to an embodiment of this application. In this embodiment, the driving data of the power semiconductor device is dynamically sampled during vehicle operation, and decision-making reasoning based on an AI diagnostic model is performed on a cloud platform. The data acquisition process in this embodiment requires the power semiconductor device to be in operation, specifically, the entire power semiconductor device must be in a high-frequency switching mode. For example, key data is sampled during the vehicle's operation.

[0322] Key data sampling is performed during the car charging, starting, and turning off stages, mainly including the following two phases: First, the collected data is transmitted to a cloud platform, where a big data-based AI diagnostic model is used for decision-making and reasoning to obtain the inference results. Second, the cloud platform sends the inference results to the vehicle terminal device, which performs data preprocessing and compression and provides emergency response functions in case of network outage. In this embodiment, data collection requires the power semiconductor to be under static conditions, i.e., the power semiconductor device is not in a low-frequency switching mode, such as during car charging, starting, and turning off stages, to perform the above key data sampling.

[0323] As illustrated by the foregoing examples, the cloud and / or edge devices in this application embodiment possess the training and inference capabilities of AI diagnostic models. This application embodiment uses AI diagnostic models to determine data characteristics and score the anomalies of these characteristics, identifying motor drives with excessive anomalies to achieve early prediction of motor drive failures. This early diagnosis of failures in the automotive electric drive system caused by motor drives and the provision of warnings allow users to perform maintenance before a malfunction occurs, preventing breakdowns during driving. Furthermore, the edge device, despite limitations in computing power, possesses local training and decision-making inference capabilities, improving the accuracy of fault prediction models and the real-time performance and reliability of inference.

[0324] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0325] To facilitate better implementation of the above-described solutions in the embodiments of this application, related apparatus for implementing the above-described solutions is also provided below.

[0326] Please see Figure 11a As shown in the embodiment of this application, a fault early warning device 1100 for a motor driver may include: an acquisition module 1101, an inference module 1102, and an early warning module 1103, wherein...

[0327] The acquisition module is used to acquire the running state data of the motor driver during operation;

[0328] The reasoning module is used to input the running state data into a preset artificial intelligence (AI) diagnostic model, and output the first anomaly parameter of the motor driver through the AI ​​diagnostic model. The AI ​​diagnostic model is used to extract data features based on the running state data and to perform decision reasoning based on the data features.

[0329] The early warning module is used to provide fault warnings to the motor driver based on the first anomaly parameter.

[0330] In some embodiments of this application, the acquisition module is further configured to acquire static data obtained by low-frequency sampling of the motor driver when the vehicle is in a stopped state, wherein the motor driver is installed on the vehicle;

[0331] The inference module is also used to input the static data into the AI ​​diagnostic model and output the second anomaly parameter of the motor driver through the AI ​​diagnostic model;

[0332] The early warning module is also used to provide fault warnings for the motor driver based on the second anomaly parameter.

[0333] In some embodiments of this application, the vehicle is in a stopped phase, including at least one of the following:

[0334] The vehicle is in the charging stage, the vehicle is in the starting stage, and the vehicle is in the off stage.

[0335] In some embodiments of this application, the early warning module is used to dynamically update the preset abnormal conditions through the AI ​​diagnostic model to obtain the updated abnormal conditions; when the first abnormality parameter meets the updated abnormal conditions, a fault warning is given to the motor driver.

[0336] In some embodiments of this application, such as Figure 11bAs shown, the device further includes: a data processing module 1104, used to clean the running state data after the acquisition module acquires the running state data of the motor driver during operation, and obtain cleaned running state data;

[0337] The inference module is used to input the cleaned running data into the AI ​​diagnostic model.

[0338] In some embodiments of this application, the device is specifically a cloud server; such as Figure 11c As shown, the device further includes: a training module 1105.

[0339] The acquisition module is also used to acquire training sample data of multiple vehicles of the same type and training sample data of multiple vehicles of different types, and the motor driver is installed on the vehicle;

[0340] The training module is used to train a model based on a fault warning task using training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types, so as to obtain the preset AI diagnostic model.

[0341] In some embodiments of this application, the device is specifically a cloud server; the device further includes: a training module.

[0342] The acquisition module is also used to acquire training sample data of one vehicle or training sample data of multiple vehicles of the same type, and the motor driver is installed on the vehicle.

[0343] The training module is used to train a model based on a fault warning task using training sample data from one vehicle or training sample data from multiple vehicles of the same type, so as to obtain the preset AI diagnostic model.

[0344] In some embodiments of this application, the device is specifically a cloud server; the early warning module is used to generate fault early warning information based on the first anomaly parameter; and send the fault early warning information to the vehicle terminal device.

[0345] In some embodiments of this application, the device is specifically a vehicle terminal device;

[0346] When the cloud server sends the AI ​​diagnostic model to the vehicle terminal device, the acquisition module is also used to receive the AI ​​diagnostic model sent by the cloud server.

[0347] In some embodiments of this application, the device is specifically a cloud server; the device further includes: a training module.

[0348] The acquisition module is also used to acquire training sample data of multiple different types of vehicles, or training sample data of one vehicle, or training sample data of multiple vehicles of the same type.

[0349] The training module is used to extract data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations; and the pre-trained model is sent to the vehicle terminal device.

[0350] In some embodiments of this application, the device is specifically a vehicle terminal device; the device further includes: a training module.

[0351] The training module is used to receive a pre-trained model from a cloud server, wherein the pre-training module represents the data association relationship obtained by the cloud server after extracting training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type; and performs model parameter optimization processing on the pre-trained model according to preset label data to obtain the preset AI diagnostic model.

[0352] In some embodiments of this application, the runtime data includes: dynamic data collected by default during the operation of the motor driver; or,

[0353] The operational data includes dynamic data collected via the CAN bus of the controller area network.

[0354] In some embodiments of this application, the operational data includes: dynamic data obtained by high-frequency sampling when the vehicle is in operation, and the motor driver is installed on the vehicle.

[0355] In some embodiments of this application, the operational data includes at least one of the following: parameters of the motor driver, parameters of the cooling system, parameters of the motor, and parameters of the battery.

[0356] In some embodiments of this application, the parameters of the motor driver include at least one of the following: DC bus voltage, DC bus current, low-voltage power supply voltage, vehicle driving status command, vehicle status, motor driver temperature, rotor position electrical angle, effective value of motor driver three-phase current, motor driver three-phase current, sampled value of motor driver three-phase current, and motor driver derating status; and / or,

[0357] The parameters of the cooling system include at least one of the following: coolant flow rate, coolant temperature, oil pump target speed, oil pump status, oil pump actual speed, oil pump power supply voltage, oil pump power semiconductor device temperature, and oil pump current; and / or,

[0358] The parameters of the motor include at least one of the following: electromagnetic frequency, motor operating mode command, motor controller operating status, motor target torque command, motor current torque, motor target speed command, motor current speed, motor temperature, motor direct-axis voltage, motor direct-axis current setpoint, motor direct-axis current feedback value, motor quadrature-axis voltage, motor quadrature-axis current setpoint, motor quadrature-axis current feedback value, and motor current verification torque; and / or,

[0359] The parameters of the battery include at least one of the following: rated output voltage, battery capacity, and maximum output current.

[0360] As illustrated by the foregoing embodiments, the process begins by acquiring operational status data of the motor drive during operation. This operational status data is then input into a preset AI diagnostic model, which outputs a first anomaly parameter for the motor drive. The AI ​​diagnostic model extracts data features from the operational status data and performs decision-making inference based on these features. A fault warning is then issued for the motor drive based on the first anomaly parameter. Since the vehicle is in operation when the motor drive is running in this embodiment, a fault warning can be issued for the motor drive based on its operational status data. This operational status data, collected during operation, reflects the actual operating state of the motor drive. By issuing a fault warning based on this operational status data, early prediction of motor drive failure can be achieved, thus enabling early warning of motor drive failures.

[0361] This application embodiment also provides a fault early warning device for a motor driver, the device comprising:

[0362] An acquisition module is used to acquire static data obtained by low-frequency sampling of the motor driver when the vehicle is in a stopped state, wherein the motor driver is installed on the vehicle;

[0363] The inference module is used to input the static data into the AI ​​diagnostic model and output the second anomaly parameter of the motor driver through the AI ​​diagnostic model;

[0364] The early warning module is used to provide fault warnings to the motor driver based on the second anomaly parameter.

[0365] In one possible implementation, the vehicle is in a stopped phase, including at least one of the following:

[0366] The vehicle is in the charging stage, the vehicle is in the starting stage, and the vehicle is in the off stage.

[0367] In one possible implementation, the early warning module is used to dynamically update the preset abnormal conditions through the AI ​​diagnostic model to obtain the updated abnormal conditions; when the second abnormality parameter meets the updated abnormal conditions, a fault warning is issued to the motor driver.

[0368] In one possible implementation, the device further includes: a data processing module, used to clean the static data obtained by the acquisition module from low-frequency sampling of the motor driver when the vehicle is in a stopped state, to obtain cleaned static data.

[0369] The inference module is used to input the cleaned static data into the AI ​​diagnostic model.

[0370] In one possible implementation, the device is specifically a cloud server; the device further includes: a training module.

[0371] The acquisition module is also used to acquire training sample data of multiple vehicles of the same type and training sample data of multiple vehicles of different types, and the motor driver is installed on the vehicle;

[0372] The training module is used to train a model based on a fault warning task using training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types, so as to obtain the preset AI diagnostic model.

[0373] In one possible implementation, the device is specifically a cloud server; the device further includes: a training module.

[0374] The acquisition module is also used to acquire training sample data of one vehicle or training sample data of multiple vehicles of the same type, and the motor driver is installed on the vehicle.

[0375] The training module is used to train a model based on a fault warning task using training sample data from one vehicle or training sample data from multiple vehicles of the same type, so as to obtain the preset AI diagnostic model.

[0376] In one possible implementation, the device is specifically a cloud server; the early warning module is used to generate fault early warning information based on the second anomaly parameter; and send the fault early warning information to the vehicle terminal device.

[0377] In one possible implementation, the device is specifically a vehicle terminal device;

[0378] When the cloud server sends the AI ​​diagnostic model to the vehicle terminal device, the acquisition module is also used to receive the AI ​​diagnostic model sent by the cloud server.

[0379] In one possible implementation, the device is specifically a cloud server; the device further includes: a training module.

[0380] The acquisition module is also used to acquire training sample data of multiple different types of vehicles, or training sample data of one vehicle, or training sample data of multiple vehicles of the same type.

[0381] The training module is used to extract data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations; and the pre-trained model is sent to the vehicle terminal device.

[0382] In one possible implementation, the device is specifically a vehicle terminal device; the device further includes: a training module.

[0383] The training module is used to receive a pre-trained model from a cloud server, wherein the pre-training module represents the data association relationship obtained by the cloud server after extracting training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type; and performs model parameter optimization processing on the pre-trained model according to preset label data to obtain the preset AI diagnostic model.

[0384] It should be noted that the information interaction and execution process between the modules / units of the above-mentioned device are based on the same concept as the method embodiments of this application, and the resulting technical effects are the same as those of the method embodiments of this application. For details, please refer to the description in the method embodiments shown above in this application, and will not be repeated here.

[0385] This application also provides a computer storage medium storing a program that performs some or all of the steps described in the above method embodiments.

[0386] The following describes another fault warning device for motor drives provided in the embodiments of this application. Please refer to [link to relevant documentation]. Figure 12 As shown, the fault warning device 1200 for a motor drive includes:

[0387] Receiver 1201, transmitter 1202, processor 1203, and memory 1204 (wherein the number of processors 1203 in the fault warning device 1200 for the motor driver can be one or more, Figure 12 (Taking a processor as an example). In some embodiments of this application, the receiver 1201, transmitter 1202, processor 1203, and memory 1204 can be connected via a bus or other means, wherein, Figure 12 Taking the example of a connection between China and Israel via a bus.

[0388] Memory 1204 may include read-only memory and random access memory, and provides instructions and data to processor 1203. A portion of memory 1204 may also include non-volatile random access memory (NVRAM). Memory 1204 stores operating system and operation instructions, executable modules or data structures, or subsets thereof, or extended sets thereof, wherein the operation instructions may include various operation instructions for implementing various operations. The operating system may include various system programs for implementing various basic business functions and handling hardware-based tasks.

[0389] Processor 1203 controls the operation of the fault warning device for the motor driver. Processor 1203 can also be referred to as a central processing unit (CPU). In specific applications, the various components of the fault warning device for the motor driver are coupled together through a bus system. This bus system includes not only a data bus but also a power bus, control bus, and status signal bus. However, for clarity, all buses are referred to as the bus system in the diagram.

[0390] The methods disclosed in the embodiments of this application can be applied to or implemented by the processor 1203. The processor 1203 can be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in the processor 1203 or by instructions in the form of software. The processor 1203 can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 1204. Processor 1203 reads the information in memory 1204 and completes the steps of the above method in conjunction with its hardware.

[0391] The receiver 1201 can be used to receive input digital or character information and generate signal inputs related to the settings and function control of the fault warning device for the motor drive. The transmitter 1202 may include a display device such as a display screen and can be used to output digital or character information through an external interface.

[0392] In this embodiment, the processor 1203 is configured to perform the following steps Figures 1 to 6 The steps are shown.

[0393] In another possible design, when the fault warning device for the motor driver is a chip within the terminal, the chip includes a processing unit and a communication unit. The processing unit may be, for example, a processor, and the communication unit may be, for example, an input / output interface, pins, or circuits. The processing unit can execute computer-executable instructions stored in a storage unit to cause the chip within the terminal to perform any of the methods described in the first aspect above. Optionally, the storage unit can be a storage unit within the chip, such as a register or cache. Alternatively, the storage unit can be a storage unit located outside the chip within the terminal, such as read-only memory (ROM) or other types of static storage devices capable of storing static information and instructions, such as random access memory (RAM).

[0394] The processor mentioned above can be a general-purpose central processing unit, a microprocessor, an ASIC, or one or more integrated circuits used to control the execution of a program in the first aspect of the method.

[0395] It should also be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. In addition, in the device embodiment drawings provided in this application, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines.

[0396] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware, or it can be implemented by special-purpose hardware including application-specific integrated circuits, special-purpose CPUs, special-purpose memory, special-purpose components, etc. Generally, any function performed by a computer program can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be diverse, such as analog circuits, digital circuits, or special-purpose circuits. However, for this application, software program implementation is more often the preferred implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium, such as a computer floppy disk, USB flash drive, mobile hard disk, ROM, RAM, magnetic disk, or optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0397] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product.

[0398] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

Claims

1. A fault early warning method for a motor driver, characterized in that, include: Obtain the running state data of the motor driver during operation; The operating state data is input into a preset artificial intelligence (AI) diagnostic model, and the AI ​​diagnostic model outputs the first anomaly parameter of the motor driver. The AI ​​diagnostic model is used to extract data features based on the operating state data and to make decision-making inferences based on the data features. The motor driver is given a fault warning based on the first anomaly parameter. Static data obtained by low-frequency sampling of the motor driver when the vehicle is stationary, wherein the motor driver is installed on the vehicle; The static data is input into the AI ​​diagnostic model, and the AI ​​diagnostic model outputs the second anomaly parameter of the motor driver. The motor driver is given a fault warning based on the second anomaly parameter.

2. The method according to claim 1, characterized in that, The vehicle being in a stopped state includes at least one of the following: The vehicle is in the charging stage, the vehicle is in the starting stage, and the vehicle is in the off stage.

3. The method according to any one of claims 1 to 2, characterized in that, The step of providing fault warning for the motor driver based on the first anomaly parameter includes: The AI ​​diagnostic model dynamically updates the preset abnormal conditions to obtain the updated abnormal conditions. When the first anomaly parameter meets the updated anomaly condition, a fault warning is issued for the motor driver.

4. The method according to any one of claims 1 to 2, characterized in that, After acquiring the running state data of the motor driver during operation, the method further includes: The running data is cleaned to obtain cleaned running data; The step of inputting the runtime data into a preset artificial intelligence (AI) diagnostic model includes: The cleaned and processed operational data is input into the AI ​​diagnostic model.

5. The method according to any one of claims 1 to 2, characterized in that, The method further includes: The cloud server acquires training sample data from multiple vehicles of the same type and multiple vehicles of different types, and the motor driver is installed on the vehicle; The cloud server uses training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types to train a model based on a fault warning task, so as to obtain the preset AI diagnostic model.

6. The method according to any one of claims 1 to 2, characterized in that, The method further includes: The cloud server acquires training sample data for one vehicle or training sample data for multiple vehicles of the same type, and the motor driver is installed on the vehicle. The cloud server uses training sample data from one vehicle or training sample data from multiple vehicles of the same type to train a model based on a fault warning task, so as to obtain the preset AI diagnostic model.

7. The method according to any one of claims 1 to 2, characterized in that, When the cloud server outputs the first anomaly parameter of the motor driver through the AI ​​diagnostic model, the step of providing a fault warning for the motor driver based on the first anomaly parameter includes: The cloud server generates fault warning information based on the first anomaly parameter. The cloud server sends the fault warning information to the vehicle terminal equipment.

8. The method according to any one of claims 1 to 2, characterized in that, When the cloud server sends the AI ​​diagnostic model to the vehicle terminal device, the method further includes: the vehicle terminal device receiving the AI ​​diagnostic model sent by the cloud server; The step of inputting the operational data into a preset artificial intelligence (AI) diagnostic model and outputting the first anomaly parameter of the motor driver through the AI ​​diagnostic model includes: The vehicle terminal device inputs the operating status data into the AI ​​diagnostic model, and outputs the first anomaly parameter of the motor driver through the AI ​​diagnostic model.

9. The method according to any one of claims 1 to 2, characterized in that, The method further includes: The cloud server acquires training sample data from multiple different types of vehicles, or training sample data from a single vehicle, or training sample data from multiple vehicles of the same type. The cloud server extracts data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations. The cloud server sends the pre-trained model to the vehicle terminal device.

10. The method according to claim 9, characterized in that, The method further includes: The vehicle terminal device receives a pre-trained model from the cloud server; The vehicle terminal device performs model parameter optimization on the pre-trained model based on preset label data to obtain the preset AI diagnostic model.

11. The method according to any one of claims 1 to 2, characterized in that, The operational data includes: dynamic data collected by default during the operation of the motor driver; or, The operational data includes dynamic data collected via the CAN bus of the controller area network.

12. The method according to any one of claims 1 to 2, characterized in that, The operational data includes dynamic data obtained by high-frequency sampling when the vehicle is in operation, and the motor driver is installed on the vehicle.

13. The method according to any one of claims 1 to 2, characterized in that, The operational data includes at least one of the following: parameters of the motor driver, parameters of the cooling system, parameters of the motor, and parameters of the battery.

14. The method according to claim 13, characterized in that, The parameters of the motor driver include at least one of the following: DC bus voltage, DC bus current, low-voltage power supply voltage, vehicle driving status command, vehicle status, motor driver temperature, rotor position electrical angle, effective value of motor driver three-phase current, motor driver three-phase current, sampled value of motor driver three-phase current, and motor driver derating status; and / or, The parameters of the cooling system include at least one of the following: coolant flow rate, coolant temperature, oil pump target speed, oil pump status, oil pump actual speed, oil pump power supply voltage, oil pump power semiconductor device temperature, and oil pump current; and / or, The parameters of the motor include at least one of the following: electromagnetic frequency, motor operating mode command, motor controller operating status, motor target torque command, motor current torque, motor target speed command, motor current speed, motor temperature, motor direct-axis voltage, motor direct-axis current setpoint, motor direct-axis current feedback value, motor quadrature-axis voltage, motor quadrature-axis current setpoint, motor quadrature-axis current feedback value, and motor current verification torque; and / or, The parameters of the battery include at least one of the following: rated output voltage, battery capacity, and maximum output current.

15. A fault early warning device for a motor driver, characterized in that, include: The acquisition module is used to acquire the running state data of the motor driver during operation; The reasoning module is used to input the running state data into a preset artificial intelligence (AI) diagnostic model, and output the first anomaly parameter of the motor driver through the AI ​​diagnostic model. The AI ​​diagnostic model is used to extract data features based on the running state data and to perform decision reasoning based on the data features. The early warning module is used to provide fault warnings to the motor driver based on the first anomaly parameter; The acquisition module is further configured to acquire static data obtained by low-frequency sampling of the motor driver when the vehicle is in a stopped state, wherein the motor driver is installed on the vehicle; The inference module is also used to input the static data into the AI ​​diagnostic model and output the second anomaly parameter of the motor driver through the AI ​​diagnostic model; The early warning module is also used to provide fault warnings for the motor driver based on the second anomaly parameter.

16. The apparatus according to claim 15, characterized in that, The vehicle being in a stopped state includes at least one of the following: The vehicle is in the charging stage, the vehicle is in the starting stage, and the vehicle is in the off stage.

17. The apparatus according to any one of claims 15 to 16, characterized in that, The early warning module is used to dynamically update the preset abnormal conditions through the AI ​​diagnostic model to obtain the updated abnormal conditions; when the first abnormality parameter meets the updated abnormal conditions, it provides a fault warning for the motor driver.

18. The apparatus according to any one of claims 15 to 16, characterized in that, The device further includes: a data processing module, used to clean the running state data of the motor driver after the acquisition module acquires the running state data during the operation of the motor driver, and obtain cleaned running state data. The inference module is used to input the cleaned running data into the AI ​​diagnostic model.

19. The apparatus according to any one of claims 15 to 16, characterized in that, The device is specifically a cloud server; the device also includes a training module. The acquisition module is also used to acquire training sample data of multiple vehicles of the same type and training sample data of multiple vehicles of different types, and the motor driver is installed on the vehicle; The training module is used to train a model based on a fault warning task using training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types, so as to obtain the preset AI diagnostic model.

20. The apparatus according to any one of claims 15 to 16, characterized in that, The device is specifically a cloud server; the device also includes a training module. The acquisition module is also used to acquire training sample data of one vehicle or training sample data of multiple vehicles of the same type, and the motor driver is installed on the vehicle. The training module is used to train a model based on a fault warning task using training sample data from one vehicle or training sample data from multiple vehicles of the same type, so as to obtain the preset AI diagnostic model.

21. The apparatus according to any one of claims 15 to 16, characterized in that, The device is specifically a cloud server; the early warning module is used to generate fault early warning information based on the first anomaly parameter and send the fault early warning information to the vehicle terminal device.

22. The apparatus according to any one of claims 15 to 16, characterized in that, The device is specifically a vehicle terminal equipment; When the cloud server sends the AI ​​diagnostic model to the vehicle terminal device, the acquisition module is also used to receive the AI ​​diagnostic model sent by the cloud server.

23. The apparatus according to any one of claims 15 to 16, characterized in that, The device is specifically a cloud server; the device also includes a training module. The acquisition module is also used to acquire training sample data of multiple different types of vehicles, or training sample data of one vehicle, or training sample data of multiple vehicles of the same type. The training module is used to extract data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations; and the pre-trained model is sent to the vehicle terminal device.

24. The apparatus according to any one of claims 15 to 16, characterized in that, The device is specifically a vehicle terminal device; the device also includes a training module. The training module is used to receive a pre-trained model from a cloud server, wherein the pre-trained model represents the data association relationship obtained by the cloud server after extracting training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type; and performs model parameter optimization processing on the pre-trained model according to preset label data to obtain the preset AI diagnostic model.

25. The apparatus according to any one of claims 15 to 16, characterized in that, The operational data includes: dynamic data collected by default during the operation of the motor driver; or, The operational data includes dynamic data collected via the CAN bus of the controller area network.

26. The apparatus according to any one of claims 15 to 16, characterized in that, The operational data includes dynamic data obtained by high-frequency sampling when the vehicle is in operation, and the motor driver is installed on the vehicle.

27. The apparatus according to any one of claims 15 to 16, characterized in that, The operational data includes at least one of the following: parameters of the motor driver, parameters of the cooling system, parameters of the motor, and parameters of the battery.

28. The apparatus according to claim 27, characterized in that, The parameters of the motor driver include at least one of the following: DC bus voltage, DC bus current, low-voltage power supply voltage, vehicle driving status command, vehicle status, motor driver temperature, rotor position electrical angle, effective value of motor driver three-phase current, motor driver three-phase current, sampled value of motor driver three-phase current, and motor driver derating status; and / or, The parameters of the cooling system include at least one of the following: coolant flow rate, coolant temperature, oil pump target speed, oil pump status, oil pump actual speed, oil pump power supply voltage, oil pump power semiconductor device temperature, and oil pump current; and / or, The parameters of the motor include at least one of the following: electromagnetic frequency, motor operating mode command, motor controller operating status, motor target torque command, motor current torque, motor target speed command, motor current speed, motor temperature, motor direct-axis voltage, motor direct-axis current setpoint, motor direct-axis current feedback value, motor quadrature-axis voltage, motor quadrature-axis current setpoint, motor quadrature-axis current feedback value, and motor current verification torque; and / or, The parameters of the battery include at least one of the following: rated output voltage, battery capacity, and maximum output current.

29. A fault early warning method for a motor driver, characterized in that, The method includes: Static data obtained by low-frequency sampling of the motor driver when the vehicle is stationary, wherein the motor driver is installed on the vehicle; The static data is input into a preset artificial intelligence (AI) diagnostic model, and the AI ​​diagnostic model outputs the second anomaly parameter of the motor driver. The motor driver is given a fault warning based on the second anomaly parameter.

30. The method according to claim 29, characterized in that, The vehicle being in a stopped state includes at least one of the following: The vehicle is in the charging stage, the vehicle is in the starting stage, and the vehicle is in the off stage.

31. The method according to claim 29 or 30, characterized in that, The step of providing fault warning for the motor driver based on the second anomaly parameter includes: The AI ​​diagnostic model dynamically updates the preset abnormal conditions to obtain the updated abnormal conditions. When the second anomaly parameter meets the updated anomaly condition, a fault warning is issued for the motor driver.

32. The method according to any one of claims 29 to 30, characterized in that, After acquiring static data obtained by low-frequency sampling of the motor driver when the vehicle is stationary, the method further includes: The static data is cleaned to obtain cleaned static data; The step of inputting the static data into a preset artificial intelligence (AI) diagnostic model includes: The static data after the cleaning process is input into the AI ​​diagnostic model.

33. The method according to any one of claims 29 to 30, characterized in that, The method further includes: The cloud server acquires training sample data from multiple vehicles of the same type and multiple vehicles of different types, and the motor driver is installed on the vehicle; The cloud server uses training sample data from multiple vehicles of the same type and training sample data from multiple vehicles of different types to train a model based on a fault warning task, so as to obtain the preset AI diagnostic model.

34. The method according to any one of claims 29 to 30, characterized in that, The method further includes: The cloud server acquires training sample data for one vehicle or training sample data for multiple vehicles of the same type, and the motor driver is installed on the vehicle. The cloud server uses training sample data from one vehicle or training sample data from multiple vehicles of the same type to train a model based on a fault warning task, so as to obtain the preset AI diagnostic model.

35. The method according to any one of claims 29 to 30, characterized in that, When the cloud server outputs the first anomaly parameter of the motor driver through the AI ​​diagnostic model, the step of providing a fault warning for the motor driver based on the second anomaly parameter includes: The cloud server generates fault warning information based on the second anomaly parameter; The cloud server sends the fault warning information to the vehicle terminal equipment.

36. The method according to any one of claims 29 to 30, characterized in that, When the cloud server sends the AI ​​diagnostic model to the vehicle terminal device, the method further includes: the vehicle terminal device receiving the AI ​​diagnostic model sent by the cloud server; The step of inputting the static data into a preset artificial intelligence (AI) diagnostic model and outputting the second anomaly parameter of the motor driver through the AI ​​diagnostic model includes: The vehicle terminal device inputs the static data into the AI ​​diagnostic model, and outputs the second anomaly parameter of the motor driver through the AI ​​diagnostic model.

37. The method according to any one of claims 29 to 30, characterized in that, The method further includes: The cloud server acquires training sample data from multiple different types of vehicles, or training sample data from a single vehicle, or training sample data from multiple vehicles of the same type. The cloud server extracts data associations from training sample data of multiple different types of vehicles, or training sample data of a single vehicle, or training sample data of multiple vehicles of the same type, to obtain a pre-trained model. The pre-trained model is used to represent the extracted data associations. The cloud server sends the pre-trained model to the vehicle terminal device.

38. The method according to claim 37, characterized in that, The method further includes: The vehicle terminal device receives a pre-trained model from the cloud server; The vehicle terminal device performs model parameter optimization on the pre-trained model based on preset label data to obtain the preset AI diagnostic model.

39. A fault early warning device for a motor driver, characterized in that, The device includes: An acquisition module is used to acquire static data obtained by low-frequency sampling of the motor driver when the vehicle is in a stopped state, wherein the motor driver is installed on the vehicle; The inference module is used to input the static data into a preset artificial intelligence (AI) diagnostic model and output the second anomaly parameter of the motor driver through the AI ​​diagnostic model. The early warning module is used to provide fault warnings to the motor driver based on the second anomaly parameter.

40. A fault early warning device for a motor drive, characterized in that, include: Memory, which stores executable program instructions; and, A processor, the processor being coupled to the memory, reads and executes instructions in the memory to cause the apparatus to perform the method as claimed in any one of claims 1 to 14, or 29 to 38.

41. The apparatus according to claim 40, characterized in that, The device is a cloud server, a chip in a cloud server, a vehicle terminal device, or a chip in a vehicle terminal device.

42. A computer-readable storage medium comprising instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 14, or 29 to 38.

43. A computer program product comprising instructions that, when run on a computer, cause the computer to perform the method as described in any one of claims 1 to 14, or 29 to 38.

Images