Vehicle-mounted charger fault self-diagnosis system, method and device based on edge calculation

Abstract

The invention provides a vehicle-mounted charger fault self-diagnosis system, method and device based on edge calculation, and relates to the technical field of new energy automobiles, the system comprises a data acquisition module, an edge calculation module and a fault prediction module; wherein the data acquisition module is used for acquiring a vehicle-mounted charger data set in the operation process of the vehicle-mounted charger; the edge calculation module is used for carrying out scheduling management on the data of the vehicle-mounted charger, and distributing calculation tasks of the data of the vehicle-mounted charger according to the priority of the data and the use condition of calculation resources so as to determine a fault diagnosis result locally on the vehicle-mounted charger; the fault prediction module comprises a grey prediction model, and is used for combining the grey prediction model with a Markov chain, carrying out fault prediction processing on the vehicle-mounted charger data, and determining a fault prediction result. According to the invention, the accuracy of fault self-diagnosis can be obviously improved.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy vehicles, and in particular, to a vehicle-mounted charger fault self-diagnosis system, method and device based on edge computing. Background Art

[0002] With the rapid development of new energy vehicles, as a key component, the reliability and stability of the vehicle-mounted charger are crucial. At present, related technologies propose that the traditional fault diagnosis method of the vehicle-mounted charger mainly relies on a centralized data processing center, and data needs to be uploaded to the cloud for analysis and diagnosis. This solution has problems such as high data transmission delay, strong network dependence, and high privacy and security risks. Therefore, it is difficult to meet the real-time fault diagnosis requirements of the vehicle-mounted charger. At the same time, the existing diagnosis methods have a low recognition accuracy for complex faults, and thus cannot effectively adapt to the operating states of the vehicle-mounted charger under different working conditions. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a vehicle-mounted charger fault self-diagnosis system, method and device based on edge computing, which can significantly improve the accuracy of fault self-diagnosis.

[0004] In a first aspect, an embodiment of the present invention provides a vehicle-mounted charger fault self-diagnosis system based on edge computing. The system includes: a data acquisition module, an edge computing module, and a fault prediction module; wherein, the data acquisition module is used to collect the vehicle-mounted charger data set during the operation of the vehicle-mounted charger, and the vehicle-mounted charger data set includes: input voltage, input current, output voltage, output current, charging temperature, charging time, switch device status information, and insulation resistance information; the edge computing module is used to schedule and manage the vehicle-mounted charger data, and allocate the calculation tasks of the vehicle-mounted charger data according to the priority of the data and the usage of computing resources, so as to determine the fault diagnosis result locally on the vehicle-mounted charger; the fault prediction module includes: a grey prediction model, and the fault prediction module is used to combine the grey prediction model with a Markov chain to perform fault prediction processing on the vehicle-mounted charger data and determine the fault prediction result.

[0005] In an implementation manner, the system further includes: a data preprocessing module; wherein, the data preprocessing module includes: an adaptive filtering model, and the data preprocessing module is used to perform data cleaning processing on the vehicle-mounted charger data set to remove noise and outliers in the data, and perform data filtering processing and normalization processing on the cleaned vehicle-mounted charger data set based on a preset adaptive filtering model to determine the target input data, so as to send the target input data to the edge computing module.

[0006] In one embodiment, the edge computing module further includes: a fault diagnosis module; wherein, the fault diagnosis module includes: a convolutional neural network and a long short-term memory network based on deep learning, and the fault diagnosis module is used to perform real-time data analysis and processing on the on-vehicle charger data according to the convolutional neural network and the long short-term memory network based on deep learning, and determine the fault diagnosis result, wherein the fault diagnosis result includes: the fault type, fault location and fault severity of the charger identified locally.

[0007] In one embodiment, the system further includes: a communication module; wherein, the communication module is used to perform data transmission and communication with the cloud server and the remote monitoring terminal, upload the fault diagnosis result and the fault prediction result to the cloud server and the remote monitoring terminal, and receive the model update instruction sent by the cloud server and the control instruction sent by the remote monitoring terminal.

[0008] In one embodiment, the system further includes: a human-computer interaction module; wherein, the human-computer interaction module includes a display screen and operation buttons, and is used to display the operating status, fault diagnosis result and fault prediction result of the on-vehicle charger to the user side through the display screen, and determine the operation instruction according to the interaction information between the user side and the operation buttons.

[0009] In a second aspect, an embodiment of the present invention provides a method for self-diagnosing faults of an on-vehicle charger based on edge computing. The method is applied to a self-diagnosing system for faults of an on-vehicle charger based on edge computing. The method includes: obtaining a set of on-vehicle charger data, and performing data preprocessing on the set of on-vehicle charger data based on an adaptive filtering model to determine target input data; performing calculation task allocation on the target input data through edge computing, and performing real-time data analysis and processing on the target input data based on a convolutional neural network and a long short-term memory network based on deep learning to determine a fault diagnosis result, wherein the fault diagnosis result includes: the fault type, fault location and fault severity of the charger identified locally; performing fault prediction processing on the on-vehicle charger data by combining a grey prediction model with a Markov chain to determine a fault prediction result, and respectively feedbacking the fault diagnosis result and the fault prediction result to the cloud server, the remote monitoring terminal and the user side.

[0010] In one embodiment, the step of performing data preprocessing on the set of on-vehicle charger data based on an adaptive filtering model to determine target input data includes: performing data cleaning processing on the set of on-vehicle charger data to remove noise and outliers in the data to determine input data; performing data filtering processing on the input data through the adaptive filtering model based on the preset adaptive filter weight coefficient, error signal and filter order of the model, and mapping the filtered input data to a preset interval through normalization processing to determine target input data.

[0011] In one embodiment, the steps of performing real-time data analysis and processing on target input data based on a deep learning convolutional neural network and a long short-term memory network to determine a fault diagnosis result include: extracting the spatial features of the target input data through a deep learning convolutional neural network to determine the spatial features of the data, and extracting the temporal features of the target input data through a long short-term memory network to determine the time series features of the data; training fault data based on the spatial features and time series features through a preset cross-entropy loss function to determine the fault diagnosis result.

[0012] In one embodiment, the steps of performing fault prediction processing on in-vehicle charger data by combining a grey prediction model and a Markov chain to determine a fault prediction result include: determining a predicted value based on the in-vehicle charger data through a grey prediction model, and predicting the possibility of a fault occurrence according to the state transition probability of the predicted value by using a Markov chain to determine the fault prediction result.

[0013] In a third aspect, an embodiment of the present invention further provides an in-vehicle charger fault self-diagnosis device based on edge computing. The device is applied to an in-vehicle charger fault self-diagnosis system based on edge computing. The device includes: a data processing unit, which acquires a set of in-vehicle charger data and performs data preprocessing on the set of in-vehicle charger data based on an adaptive filtering model to determine target input data; a fault diagnosis unit, which assigns a computing task to the target input data through edge computing and performs real-time data analysis and processing on the target input data based on a deep learning convolutional neural network and a long short-term memory network to determine a fault diagnosis result. The fault diagnosis result includes: the fault type, fault location, and fault severity of the charger identified locally; a fault prediction unit, which performs fault prediction processing on the in-vehicle charger data by combining a grey prediction model and a Markov chain to determine a fault prediction result, and feeds back the fault diagnosis result and the fault prediction result to a cloud server, a remote monitoring terminal, and a user terminal respectively.

[0014] In a fourth aspect, an embodiment of the present invention further provides a server, which includes a processor and a memory. The memory stores computer-executable instructions that can be executed by the processor. The processor executes the computer-executable instructions to implement the method according to any one of the second aspect.

[0015] In a fifth aspect, an embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions cause the processor to implement the method according to any one of the second aspect.

[0016] The embodiments of the present invention bring the following beneficial effects: An on-vehicle charger fault self-diagnosis system, method and device based on edge computing provided by an embodiment of the present invention. The system includes: a data acquisition module, an edge computing module, and a fault prediction module. Among them, the data acquisition module is used to collect an on-vehicle charger data set during the operation of the on-vehicle charger. The on-vehicle charger data set includes: input voltage, input current, output voltage, output current, charging temperature, charging time, switching device status information, and insulation resistance information. The edge computing module is used to schedule and manage the on-vehicle charger data, and allocate the calculation tasks of the on-vehicle charger data according to the priority of the data and the usage of computing resources, so as to determine the fault diagnosis result locally on the on-vehicle charger. The fault prediction module includes: a grey prediction model. The fault prediction module is used to combine the grey prediction model with a Markov chain to perform fault prediction processing on the on-vehicle charger data and determine the fault prediction result. The embodiment of the present invention can deploy an edge computing module to realize local processing and fault diagnosis of data, thereby reducing data transmission delay, improving the real-time performance of fault diagnosis, and meeting the real-time operation requirements of the on-vehicle charger. In addition, by adopting a fault diagnosis model that combines CNN and LSTM, and a fault prediction model that combines a grey prediction model and a Markov chain, the accuracy of fault diagnosis and prediction can be significantly improved, and various complex faults and potential faults can be effectively identified.

[0017] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention are realized and obtained by the structures specifically pointed out in the specification, claims, and drawings.

[0018] To make the above objectives, features, and advantages of the present invention more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, provides a detailed description as follows. Description of the Drawings

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0020] Figure 1 It is a schematic structural diagram of an on-vehicle charger fault self-diagnosis system based on edge computing provided by an embodiment of the present invention; Figure 2 It is a specific structural diagram of an on-vehicle charger fault self-diagnosis system based on edge computing provided by an embodiment of the present invention; Figure 3Schematic flowchart of a vehicle-mounted charger fault self-diagnosis method based on edge computing provided by an embodiment of the present invention; Figure 4 Schematic structural diagram of a vehicle-mounted charger fault self-diagnosis device based on edge computing provided by an embodiment of the present invention; Figure 5 Schematic structural diagram of a server provided by an embodiment of the present invention. Detailed implementation manners

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0022] Currently, with the rapid development of new energy vehicles, as a key component, the reliability and stability of vehicle-mounted chargers are crucial. Related technologies propose that the traditional vehicle-mounted charger fault diagnosis method mainly relies on a centralized data processing center, and data needs to be uploaded to the cloud for analysis and diagnosis. This solution has problems such as high data transmission delay, strong network dependence, and high privacy and security risks. Therefore, it is difficult to meet the requirements of real-time fault diagnosis of vehicle-mounted chargers. At the same time, the existing diagnosis methods have a low recognition accuracy for complex faults, and thus cannot effectively adapt to the operating states of vehicle-mounted chargers under different working conditions. Based on this, the vehicle-mounted charger fault self-diagnosis system, method, and device based on edge computing provided by the embodiments of the present invention can deploy an edge computing module to realize local processing and fault diagnosis of data, thereby reducing data transmission delay, improving the real-time performance of fault diagnosis, and meeting the requirements of real-time operation of vehicle-mounted chargers. In addition, by adopting a fault diagnosis model that combines CNN and LSTM, and a fault prediction model that combines a grey prediction model and a Markov chain, the accuracy of fault diagnosis and prediction can be significantly improved, and various complex faults and potential faults can be effectively identified.

[0023] To facilitate the understanding of this embodiment, first, a vehicle-mounted charger fault self-diagnosis method based on edge computing disclosed in the embodiments of the present invention will be introduced in detail. This method is applied to a vehicle-mounted charger fault self-diagnosis system based on edge computing. To facilitate the understanding of the vehicle-mounted charger fault self-diagnosis system based on edge computing, the embodiments of the present invention provide a schematic structural diagram of a vehicle-mounted charger fault self-diagnosis based on edge computing, as Figure 1 shown. The system includes: a data acquisition module, an edge computing module, and a fault prediction module.

[0024] The data acquisition module is used to collect the on-board charger data set during the operation of the on-board charger. Among them, the on-board charger data set includes: input voltage, input current, output voltage, output current, charging temperature, charging time, switch device status information, and insulation resistance information. Specifically, the sensors of the data acquisition module are selected as follows: Current detection: Use Honeywell SL353 series Hall effect sensors (accuracy ±0.5%FS, response time <1μs), support differential current measurement, and are suitable for current monitoring of high-frequency switching devices.

[0025] Temperature detection: Deploy PT100 temperature sensors (accuracy ±0.1°C), cooperate with LM-PT100 acquisition modules (RS-485 interface, support three-wire / four-wire compensation), and realize multi-point temperature monitoring of the charger's power devices and heat sinks.

[0026] Insulation detection: Integrate a Hipot HIOKI 3193 insulation resistance tester (test voltage 500VDC, range 0.1MΩ - 1GΩ), and periodically detect the insulation resistance of the high-voltage circuit to the ground.

[0027] Signal conditioning: Use an AD7606 16-bit synchronous sampling ADC chip (8 channels, 200kSPS) to achieve high-precision digital conversion of analog signals.

[0028] The edge computing module is used to schedule and manage the on-board charger data. According to the priority of the data and the usage of computing resources, it allocates the computing tasks of the on-board charger data to determine the fault diagnosis result locally on the on-board charger. Specifically, the edge computing module can be built based on an embedded processor, equipped with a lightweight operating system and computing framework, receive the data transmitted by the data preprocessing module, and use the pre-trained fault diagnosis model and fault prediction model for data processing and analysis. It has task scheduling and resource management functions, and can reasonably allocate computing tasks according to the priority of the data and the usage of computing resources. Among them, the hardware platform of the edge computing module: Based on the NVIDIA Jetson Nano development kit (4GB memory, 128-core GPU), equipped with the Ubuntu 18.04 operating system and the TensorFlowLite framework, supporting real-time inference; the resource management of the edge computing module: Use cgroups to limit the memory occupancy of the fault diagnosis model (about 500MB memory) and the prediction model (about 200MB memory); in addition, adopt the dynamic voltage and frequency scaling (DVFS) technology, which can automatically adjust the CPU frequency (528MHz - 1.43GHz) according to the load.

[0029] The edge computing module further includes: a fault diagnosis module, which includes: a convolutional neural network and a long short-term memory network based on deep learning. The fault diagnosis module is used to perform real-time data analysis and processing on the on-vehicle charger data according to the convolutional neural network and the long short-term memory network based on deep learning, and determine the fault diagnosis result. Among them, the fault diagnosis result includes: the fault type, fault location and fault severity of the charger identified locally. That is to say, the fault diagnosis module is responsible for analyzing and processing the data generated during the operation of the on-vehicle charger, and identifying the possible fault types, locations and severities of the charger through specific algorithms and rules.

[0030] The fault prediction module includes: a grey prediction model. The fault prediction module is used to combine the grey prediction model with a Markov chain to perform fault prediction processing on the on-vehicle charger data and determine the fault prediction result. The grey prediction model is a prediction method based on grey system theory, which is applicable to situations with incomplete or highly uncertain data. Its basic principle is to generate an accumulated sequence from the original data to transform the non-stationary sequence into a stationary sequence, and then use a differential equation model for prediction. The grey prediction model has strong adaptability, especially suitable for predicting small sample data. The future operating state of the on-vehicle charger is predicted through the fault prediction model to determine whether there are potential faults and output the fault prediction result.

[0031] See Figure 2 The specific structural schematic diagram of an on-vehicle charger fault self-diagnosis system based on edge computing shown in

[0032] The data preprocessing module includes: an adaptive filtering model. The data preprocessing module is used to perform data cleaning on the on-vehicle charger data set to remove noise and outliers in the data, and perform data filtering and normalization processing on the cleaned on-vehicle charger data set based on a preset adaptive filtering model to determine the target input data, so as to send the target input data to the edge computing module. That is to say, the data preprocessing module performs cleaning, filtering and normalization processing on the collected data, removes noise and outliers, improves the data quality, and then transmits the processed data to the edge computing module.

[0033] The communication module is used for data transmission and communication with the cloud server and the remote monitoring terminal, uploading the fault diagnosis results and fault prediction results to the cloud server and the remote monitoring terminal, and receiving the model update instructions sent by the cloud server and the control instructions sent by the remote monitoring terminal. Among them, the communication module supports multiple communication protocols, including in-vehicle Ethernet, CAN bus, and 4G / 5G mobile communication; uploading the fault diagnosis results and fault prediction information to the cloud server and the remote monitoring terminal, and at the same time receiving the model update instructions from the cloud server and the control instructions from the remote monitoring terminal. Further, the in-vehicle network of the communication module: interacts with systems such as BMS and VCU through the CAN bus (baud rate 500 kbps) to transmit the charging status and control instructions. The remote communication of the communication module: integrates the Quectel EC200U 5G module, supports NSA / SA dual mode, and conducts data interaction with the cloud server (EMQX cluster) through the MQTT protocol (QoS1) to transmit fault data and receive model updates.

[0034] The human-machine interaction module includes a display screen and operation buttons, which are used to display the operating status, fault diagnosis results, and fault prediction results of the on-vehicle charger to the user side through the display screen, and determine the operation instructions according to the interaction information between the user side and the operation buttons. The human-machine interaction module provides an intuitive and convenient operation interface for the user side, facilitating the driver to understand the operating status and fault information of the on-vehicle charger and enhancing the user experience.

[0035] In addition, the system also includes a data management module. Based on the integrated function of the graph model library and following the IEC61970 CIM standard, the data management module realizes the unified input, storage, and management of power grid models, parameters, and graphic data.

[0036] Based on Figure 1 the structural schematic diagram of the on-vehicle charger fault self-diagnosis system based on edge computing shown in Figure 2 and the specific structural schematic diagram of the on-vehicle charger fault self-diagnosis system based on edge computing shown in Figure 3 the flow schematic diagram of an on-vehicle charger fault self-diagnosis method based on edge computing shown in Step S302: Obtain the in-vehicle charger data set, and based on the adaptive filtering model, perform data preprocessing on the in-vehicle charger data set to determine the target input data. In one implementation, data cleaning processing can be performed on the in-vehicle charger data set to remove noise and outliers in the data to determine the input data. Then, through the adaptive filtering model, based on the preset adaptive filter weight coefficients, error signals, and filter order of the model, perform data filtering processing on the input data, and map the filtered input data to the preset interval through normalization processing to determine the target input data. Among them, the adaptive filtering model is:

[0037] Among them, is the filtered output signal, is the input signal, is the weight coefficient of the adaptive filter, is the error signal, is the filter order.

[0038] Normalization processing maps the data to the interval:

[0039] Among them, is the normalized data, is the original data, and are the maximum and minimum values of the original data respectively; the processed data is transmitted to the edge computing module.

[0040] Step S304: Through edge computing, perform calculation task allocation on the target input data, and based on the convolutional neural network and long short-term memory network of deep learning, perform real-time data analysis processing on the target input data to determine the fault diagnosis result. Among them, the fault diagnosis result includes: the fault type, fault location, and fault severity of the charger identified locally. In one implementation, the spatial features of the target input data can be extracted through the convolutional neural network based on deep learning to determine the spatial features of the data, and the time features of the target input data can be extracted through the long short-term memory network to determine the time series features of the data. Then, through the preset cross-entropy loss function, perform fault data training based on the spatial features and time series features to determine the fault diagnosis result.

[0041] Specifically, the fault diagnosis module adopts a fault diagnosis model that fuses the convolutional neural network (CNN) based on deep learning and the long short-term memory network (LSTM); the CNN is used to extract the spatial features of the data, and the LSTM is used to process the time series features of the data. Then, the spatial features and time series features are mapped to the fault categories to determine the number of fault categories C, and the cross-entropy loss function is used for fault training. The cross-entropy loss function is as follows:

[0042] where is the number of samples, is the number of fault categories, is the sample belonging to the category true label, is the sample belonging to the category predicted probability. Through the training of a large amount of historical fault data, the cross-entropy loss function enables the model to accurately identify various fault types of the on-vehicle charger and output the fault diagnosis results.

[0043] Step S306: By combining the grey prediction model with the Markov chain, perform fault prediction processing on the on-vehicle charger data to determine the fault prediction results, and feedback the fault diagnosis results and fault prediction results to the cloud server, remote monitoring terminal, and user terminal respectively. In one implementation, the grey prediction model can be used to determine the predicted value based on the on-vehicle charger data, and the Markov chain can be used to predict the possibility of a fault according to the state transition probability of the predicted value to determine the fault prediction result. The grey prediction model is used to predict the trend of the on-vehicle charger operating parameters, and the formula is:

[0044] where is the predicted value, and are model parameters, is the first value of the original data; the Markov chain is used to predict the possibility of a fault according to the state transition probability of the predicted value.

[0045] Furthermore, the following test cases can be used for functional verification: Overvoltage fault: Simulate the input voltage rising to 420V, and the system triggers protection within 200 ms and sends a fault code 0x123 through the CAN bus.

[0046] Overheat warning: Heat the power device to 85°C, and the system predicts that a fault may occur within the next 5 minutes and displays a red warning through the display screen.

[0047] Performance indicators: Diagnostic latency: Local processing is performed when it is less than 100 ms, and cloud collaboration is required when it is less than 500 ms; Prediction accuracy: The accuracy of predicting faults 10 minutes in advance reaches 92%.

[0048] In practical applications, Scenario 1: When the insulation resistance drops to 0.5 MΩ during the charging process, the system immediately cuts off the high-voltage circuit and uploads the fault log to the cloud through 5G; Scenario 2: When abnormal fan speed is detected during vehicle startup, the system receives the updated diagnostic model through OTA to optimize the fan fault identification logic.

[0049] In summary, the present invention can deploy an edge computing module on the vehicle-mounted side to achieve local processing and fault diagnosis of data, thereby reducing data transmission latency, improving the real-time performance of fault diagnosis, and meeting the real-time operation requirements of the on-vehicle charger. In addition, by adopting a fault diagnosis model that combines CNN and LSTM, and a fault prediction model that combines the grey prediction model and Markov chain, the accuracy of fault diagnosis and prediction can be significantly improved, and various complex faults and potential faults can be effectively identified.

[0050] Furthermore, the system also has multiple communication methods, which can realize the interconnection and interoperability between the on-vehicle charger and other systems, cloud servers, and remote monitoring terminals, facilitating remote monitoring and management, and improving the intelligent level of new energy vehicles. In addition, the human-computer interaction module provides an intuitive and convenient operation interface for the driver, facilitating the driver to understand the operating status and fault information of the on-vehicle charger, and enhancing the user experience.

[0051] For the on-vehicle charger fault self-diagnosis method based on edge computing provided in the foregoing embodiments, the embodiments of the present invention provide an on-vehicle charger fault self-diagnosis device based on edge computing. This device is applied to an on-vehicle charger fault self-diagnosis system based on edge computing. Refer to Figure 4 The structural schematic diagram of an on-vehicle charger fault self-diagnosis device based on edge computing shown. This device includes the following parts: A data processing unit 402, which acquires the on-vehicle charger data set and performs data preprocessing on the on-vehicle charger data set based on an adaptive filtering model to determine the target input data; A fault diagnosis unit 404, which assigns calculation tasks to the target input data through edge computing, and performs real-time data analysis and processing on the target input data based on the convolutional neural network and long short-term memory network of deep learning to determine the fault diagnosis result. Among them, the fault diagnosis result includes: the fault type, fault location, and fault severity of the charger identified locally; The fault prediction unit 406 combines a grey prediction model with a Markov chain to perform fault prediction processing on the on-vehicle charger data, determines the fault prediction result, and feeds back the fault diagnosis result and the fault prediction result to the cloud server, the remote monitoring terminal, and the user terminal respectively.

[0052] The above-mentioned on-vehicle charger fault self-diagnosis device based on edge computing provided by the embodiments of the present application can significantly improve the accuracy of fault self-diagnosis.

[0053] In one implementation manner, when performing the step of preprocessing the on-vehicle charger data set based on an adaptive filtering model to determine the target input data, the above-mentioned data processing unit 402 is further configured to: perform data cleaning processing on the on-vehicle charger data set to remove noise and outliers in the data and determine the input data; perform data filtering processing on the input data based on the preset adaptive filter weight coefficient, error signal, and filter order of the model through the adaptive filtering model, and map the filtered input data to a preset interval through normalization processing to determine the target input data.

[0054] In one implementation manner, when performing the step of performing real-time data analysis processing on the target input data based on a convolutional neural network and a long short-term memory network of deep learning to determine the fault diagnosis result, the above-mentioned fault diagnosis unit 404 is further configured to: extract the spatial features of the target input data through a convolutional neural network based on deep learning to determine the spatial features of the data, and extract the time features of the target input data through a long short-term memory network to determine the time series features of the data; perform fault data training based on the spatial features and the time series features through a preset cross-entropy loss function to determine the fault diagnosis result.

[0055] In one implementation manner, when performing the step of combining a grey prediction model with a Markov chain to perform fault prediction processing on the on-vehicle charger data to determine the fault prediction result, the above-mentioned fault prediction unit 406 is further configured to: determine a predicted value based on the on-vehicle charger data through the grey prediction model, and use the Markov chain to predict the possibility of a fault occurring according to the state transition probability of the predicted value to determine the fault prediction result.

[0056] The device provided by the embodiments of the present invention has the same implementation principle and the same technical effects as those of the foregoing method embodiments. For a brief description, for the parts not mentioned in the device embodiments, reference may be made to the corresponding contents in the foregoing method embodiments.

[0057] The embodiments of the present invention provide a server. Specifically, the server includes a processor and a storage device; a computer program is stored on the storage device, and when the computer program is run by the processor, it executes the method according to any one of the above-mentioned implementation manners.

[0058] Figure 5 The following is a schematic structural diagram of a server provided by an embodiment of the present invention. The server 100 includes: a processor 50, a memory 51, a bus 52, and a communication interface 53. The processor 50, the communication interface 53, and the memory 51 are connected through the bus 52. The processor 50 is configured to execute an executable module stored in the memory 51, such as a computer program.

[0059] Among them, the memory 51 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between this system network element and at least one other network element is realized through at least one communication interface 53 (which can be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. can be used.

[0060] The bus 52 may be an ISA bus, a PCI bus, an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 5 only a bidirectional arrow is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0061] Among them, the memory 51 is used to store a program. After receiving an execution instruction, the processor 50 executes the program. The method executed by the device defined by the flow process disclosed in any one of the foregoing embodiments of the present invention can be applied to the processor 50 or implemented by the processor 50.

[0062] The processor 50 may be an integrated circuit chip with the ability to process signals. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in the processor 50 or the instructions in the form of software. The above-mentioned processor 50 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP for short), an application specific integrated circuit (ASIC for short), a field-programmable gate array (FPGA for short), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present invention can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of the hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory 51, and the processor 50 reads the information in the memory 51 and combines its hardware to complete the steps of the above method.

[0063] The computer program product of the readable storage medium provided by the embodiments of the present invention includes a computer-readable storage medium storing program code, and the instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For the specific implementation, reference can be made to the foregoing method embodiments and will not be elaborated here.

[0064] If the above-described functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs.

[0065] Finally, it should be noted that the above-described embodiments are only specific implementation manners of the present invention, used to illustrate the technical solutions of the present invention, rather than limiting them. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that any technician familiar with the technical field of the present invention can still modify the technical solutions described in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims

1. An on-vehicle charger fault self-diagnosis system based on edge computing, characterized in that, The system includes: a data acquisition module, an edge computing module, and a fault prediction module; wherein, The data acquisition module is used to collect the in-vehicle charger data set during the operation of the in-vehicle charger. Among them, the in-vehicle charger data set includes: input voltage, input current, output voltage, output current, charging temperature, charging time, switch device status information, and insulation resistance information; The edge computing module is used to schedule and manage the in-vehicle charger data, and allocate the calculation tasks of the in-vehicle charger data according to the priority of the data and the usage of computing resources, so as to determine the fault diagnosis result locally on the in-vehicle charger; The fault prediction module includes: a grey prediction model. The fault prediction module is used to combine the grey prediction model with the Markov chain to perform fault prediction processing on the in-vehicle charger data and determine the fault prediction result.

2. The on-vehicle charger fault self-diagnosis system based on edge computing according to claim 1, wherein, The system further includes: a data preprocessing module; wherein, The data preprocessing module includes: an adaptive filtering model. The data preprocessing module is used to perform data cleaning processing on the in-vehicle charger data set to remove noise and outliers in the data, and based on a preset adaptive filtering model, perform data filtering processing and normalization processing on the cleaned in-vehicle charger data set to determine the target input data, so as to send the target input data to the edge computing module.

3. The on-vehicle charger fault self-diagnosis system based on edge computing according to claim 1, characterized in that The edge computing module further includes: a fault diagnosis module; wherein, The fault diagnosis module includes: a convolutional neural network and a long short-term memory network based on deep learning. The fault diagnosis module is used to perform real-time data analysis processing on the in-vehicle charger data according to the convolutional neural network and the long short-term memory network based on deep learning, and determine the fault diagnosis result. Among them, the fault diagnosis result includes: the fault type, fault location, and fault severity of the charger identified locally.

4. The on-vehicle charger fault self-diagnosis system based on edge computing according to claim 1, characterized in that, The system further includes: a communication module; wherein, The communication module is used to perform data transmission and communication with the cloud server and the remote monitoring terminal, upload the fault diagnosis result and the fault prediction result to the cloud server and the remote monitoring terminal, and receive the model update instruction sent by the cloud server and the control instruction sent by the remote monitoring terminal.

5. The on-vehicle charger fault self-diagnosis system based on edge computing according to claim 1, characterized in that, The system further includes: a human-computer interaction module; wherein, The human-computer interaction module includes a display screen and operation buttons, and is used to display the operation status of the in-vehicle charger, the fault diagnosis result, and the fault prediction result to the user side through the display screen, and determine the operation instruction according to the interaction information between the user side and the operation buttons.

6. A vehicle-mounted charger fault self-diagnosis method based on edge computing, characterized in that, The method is applied to an in-vehicle charger fault self-diagnosis system based on edge computing. The method includes: Obtain the in-vehicle charger data set, and based on the adaptive filtering model, perform data preprocessing on the in-vehicle charger data set to determine the target input data; Perform computational task allocation on the target input data through edge computing, and perform real-time data analysis and processing on the target input data based on the convolutional neural network and long short-term memory network of deep learning to determine the fault diagnosis result. Among them, the fault diagnosis result includes: the fault type, fault location, and fault severity of the charger identified locally; Through the combination of the grey prediction model and the Markov chain, perform fault prediction processing on the in-vehicle charger data, determine the fault prediction result, and feedback the fault diagnosis result and the fault prediction result to the cloud server, remote monitoring terminal, and user terminal respectively.

7. The on-vehicle charger fault self-diagnosis method based on edge computing according to claim 6, characterized in that, The steps of performing data preprocessing on the in-vehicle charger data set based on the adaptive filtering model to determine the target input data include: Perform data cleaning processing on the in-vehicle charger data set to remove noise and outliers in the data and determine the input data; Through the adaptive filtering model, based on the preset adaptive filter weight coefficients, error signals, and filter order of the model, perform data filtering processing on the input data, and map the filtered input data to the preset interval through normalization processing to determine the target input data.

8. The on-vehicle charger fault self-diagnosis method based on edge computing according to claim 6, wherein The steps of performing real-time data analysis and processing on the target input data based on the convolutional neural network and long short-term memory network of deep learning to determine the fault diagnosis result include: Extract the spatial features of the target input data through the convolutional neural network based on deep learning to determine the spatial features of the data, and extract the time features of the target input data through the long short-term memory network to determine the time series features of the data; Determine the fault diagnosis result through the preset cross-entropy loss function based on the spatial features and the time series features.

9. The on-vehicle charger fault self-diagnosis method based on edge computing according to claim 6, wherein, The steps of performing fault prediction processing on the in-vehicle charger data through the combination of the grey prediction model and the Markov chain to determine the fault prediction result include: Through the grey prediction model, based on the in-vehicle charger data, determine the predicted value, and use the Markov chain to predict the possibility of fault occurrence according to the state transition probability of the predicted value to determine the fault prediction result.

10. A vehicle-mounted charger fault self-diagnosis device based on edge computing, characterized in that, The device is applied to an in-vehicle charger fault self-diagnosis system based on edge computing. The device includes: A data processing unit that acquires the in-vehicle charger data set and performs data preprocessing on the in-vehicle charger data set based on the adaptive filtering model to determine the target input data; A fault diagnosis unit that performs computational task allocation on the target input data through edge computing, and performs real-time data analysis and processing on the target input data based on the convolutional neural network and long short-term memory network of deep learning to determine the fault diagnosis result. Among them, the fault diagnosis result includes: the fault type, fault location, and fault severity of the charger identified locally; The fault prediction unit combines a grey prediction model with a Markov chain to perform fault prediction processing on the data of the on-vehicle charger, determine the fault prediction result, and respectively feedback the fault diagnosis result and the fault prediction result to the cloud server, the remote monitoring terminal and the user terminal.

11. A server, characterized in that, It includes a processor and a memory. The memory stores computer executable instructions that can be executed by the processor. The processor executes the computer executable instructions to implement the method according to any one of claims 6 to 9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer executable instructions. When the computer executable instructions are called and executed by the processor, the computer executable instructions prompt the processor to implement the method according to any one of claims 6 to 9.

Images