New energy vehicle fault diagnosis and early warning system based on artificial intelligence

Abstract

The invention discloses a new energy automobile fault diagnosis and early warning system based on artificial intelligence, and the system employs an advanced artificial intelligence technology, and is designed according to the fault diagnosis and early warning demands of a new energy automobile. By integrating a high-precision sensor and a real-time data acquisition module, the system can comprehensively monitor key performance parameters of the vehicle; by utilizing machine learning and deep learning algorithms, the system can automatically identify a fault mode, the diagnosis accuracy is improved, and the diagnosis time is shortened; meanwhile, based on a potential fault early warning function of big data analysis, vehicle data can be monitored in real time, potential faults can be found in advance, and measures can be taken in time to prevent the faults. The system also has a strong data processing capability, can process and analyze a large amount of vehicle data in real time, and provides powerful support for fault diagnosis and early warning. Generally speaking, the system can significantly improve the safety, reliability and maintenance efficiency of the new energy vehicle, reduce the maintenance cost of a vehicle owner, and improve the user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy vehicles, and specifically to a new energy vehicle fault diagnosis and early warning system based on artificial intelligence, which aims to achieve real-time monitoring, fault diagnosis and early warning of key components of new energy vehicles by integrating artificial intelligence algorithms, thereby improving the safety and reliability of new energy vehicles. Background Art

[0002] With the rapid development of new energy vehicle technology, its fault diagnosis and early warning system has also made significant progress. At present, the fault diagnosis and early warning of new energy vehicle power systems mainly rely on on-board electronic control units (ECUs) and sensors and other devices. These devices can monitor various performance parameters of the vehicle in real time, such as battery voltage, current, temperature, motor speed, torque, etc., and automatically alarm the driver when a fault occurs. In addition, with the popularization of Internet of Things (IoT) technology, vehicle data can be uploaded to the cloud platform in real time, and artificial intelligence technology can be used for in-depth analysis and fault diagnosis, thereby realizing early warning of potential faults.

[0003] In the field of fault diagnosis of new energy vehicles, the application of artificial intelligence technology is becoming more and more extensive. Algorithms based on machine learning and deep learning can process and analyze large amounts of vehicle data, identify fault modes, and improve the accuracy of diagnosis. For example, algorithms such as support vector machines, convolutional neural networks, and long short-term memory networks perform well in fault classification and prediction. These algorithms can automatically extract features, classify faults, and adapt to different vehicle components and fault types.

[0004] Although the current new energy vehicle fault diagnosis and early warning system has achieved certain results, there are still some shortcomings:

[0005] Inaccurate fault judgment: Due to the accumulation of errors in various sensors, imperfect software algorithms, and data noise, the current fault diagnosis system is prone to misjudgment. This may lead to the wrong replacement and repair of normal parts, increasing the maintenance costs of car owners.

[0006] Long fault diagnosis time: For some complex faults, the current fault diagnosis system takes a long time to find the exact cause. This affects the user experience and may lead to further deterioration of the fault.

[0007] Imperfect early warning system: The current fault early warning system mainly provides prompts for faults that have already occurred, and lacks the early warning function for potential faults. This cannot prevent the occurrence of faults in advance, reducing the overall reliability of the system.

[0008] Limited data processing capabilities: With the popularity of new energy vehicles and the improvement of their intelligence, the amount of data generated by vehicles has increased dramatically. The current fault diagnosis and early warning system has bottlenecks in data processing capabilities, making it difficult to process and analyze large amounts of vehicle data in real time. Summary of the invention

[0009] The present invention aims to integrate modules such as data collection, data analysis, fault diagnosis, early warning release, information management and communication, and utilize artificial intelligence algorithms to conduct real-time monitoring and fault warning of key components of new energy vehicles, so as to improve the safety, reliability and maintenance efficiency of the vehicles.

[0010] To achieve the above object, the present invention adopts the following technical solutions:

[0011] A new energy vehicle fault diagnosis and early warning system based on artificial intelligence, comprising:

[0012] a) Data acquisition module, which collects driving data of key components and characteristic indicators of driving conditions of new energy vehicles in real time;

[0013] b) A data analysis module receives data transmitted by the data acquisition module and analyzes the data using an artificial intelligence algorithm;

[0014] c) Fault diagnosis module, which uses machine learning models to perform fault diagnosis based on the results of the data analysis module and outputs the fault type and fault severity;

[0015] d) Early warning module, which receives the diagnosis result of the fault diagnosis module and sends early warning information to the driver or maintenance personnel when a serious or urgent fault is detected;

[0016] e) Information management module, used to store, manage and display vehicle operation data, fault history records and warning information;

[0017] f) Communication module, which realizes data communication between modules and data transmission between the system and external devices.

[0018] Furthermore, the data acquisition module includes a plurality of sensors, specifically:

[0019] Battery voltage sensor: collects voltage data of new energy vehicle battery packs in real time;

[0020] Battery current sensor: collects current data of new energy vehicle battery packs in real time;

[0021] Motor temperature sensor: collects temperature data of new energy vehicle motors in real time;

[0022] Vehicle speed sensor: collects vehicle speed data of new energy vehicles in real time.

[0023] Furthermore, the data analysis module uses the following formula to calculate the abnormality index of the data:

[0024]

[0025] Among them, x i is the value of the ith data point, μ i is the historical average value of the data point, σ i is the historical standard deviation of the data point, and n is the number of data points; when AI exceeds the preset threshold, it is judged as abnormal.

[0026] Furthermore, the data analysis module also includes:

[0027] The data preprocessing unit is used to clean, denoise and normalize the collected data. The following formula is used for normalization:

[0028]

[0029] Among them, x i is the original data, x min and x max are the minimum and maximum values ​​in the data set, respectively, and x norm，i is the normalized data;

[0030] A feature extraction unit extracts key information that can reflect fault characteristics from the preprocessed data;

[0031] The anomaly detection unit uses the above anomaly index calculation formula to perform anomaly detection on the data after feature extraction, and outputs the abnormal data to the fault diagnosis module.

[0032] Furthermore, the machine learning model of the fault diagnosis module further includes:

[0033] Input layer, receives abnormal data, and the number of neurons in the input layer is consistent with the number of features extracted by the feature extraction unit;

[0034] The hidden layer contains multiple layers of neurons. The number of neurons in each layer is determined according to the model complexity and the size of the data set. The neurons in the hidden layer use the ReLU activation function for nonlinear transformation. The RELU function is defined as:

[0035] RELU(x)=max(0,x)

[0036] The output layer uses the softmax function to output the probability distribution of fault type and fault degree. The softmax function is defined as:

[0037]

[0038] Among them, zj is the output of the jth neuron, and K is the total number of neurons in the output layer, that is, the number of fault types.

[0039] Furthermore, the early warning module also includes:

[0040] The early warning strategy formulation unit formulates corresponding early warning strategies according to the fault type and severity, including the use of sound and light alarms, SMS notifications or APP push;

[0041] The warning information generation unit generates specific warning information according to the warning strategy, including the fault type, fault degree and recommended measures.

[0042] Furthermore, the information management module also includes:

[0043] Data storage unit, used to store vehicle operation data, fault history records and warning information. The data is stored in a structured database for easy query and analysis;

[0044] The data display unit displays the stored data to the driver or maintenance personnel in the form of charts and reports, providing intuitive vehicle status monitoring and fault analysis.

[0045] Furthermore, the communication module adopts wireless communication technology to realize data communication between modules and data transmission between the system and external devices. The data transmission adopts encryption protocol to ensure data security and privacy protection.

[0046] The present invention proposes a current fluctuation prediction method based on artificial intelligence, which has the following beneficial effects:

[0047] 1. Improve the accuracy of fault diagnosis: By introducing sensor technology and optimizing fault diagnosis algorithms, the system can more accurately identify fault modes and reduce misjudgments and missed judgments, which helps reduce the maintenance costs of car owners and improve maintenance efficiency;

[0048] 2. Shorten fault diagnosis time: Using machine learning, the system can quickly analyze vehicle data, find out the cause of the fault, and give solution suggestions. This can greatly shorten the fault diagnosis time and improve the user experience;

[0049] 3. Achieve early warning of potential faults: By establishing a potential fault early warning system based on big data and machine learning, the system can monitor vehicle data in real time, detect potential faults in advance, and take timely measures to repair them, which helps prevent the occurrence of faults and improve the overall reliability of the system;

[0050] 4. Improve data processing capabilities: Using computing platforms and distributed data processing technology, the system can process and analyze large amounts of vehicle data in real time, which helps to mine more useful information and provide more accurate data support for fault diagnosis and early warning.

[0051] In summary, the new energy vehicle fault diagnosis and early warning system based on artificial intelligence has significant beneficial effects, can overcome the shortcomings of current technology, and improve the safety, reliability and maintenance efficiency of new energy vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a schematic diagram of the system structure of a new energy vehicle fault diagnosis and early warning system based on artificial intelligence in the present invention;

[0054] Figure 2 This is an implementation flow chart of an artificial intelligence-based new energy vehicle fault diagnosis and early warning system of the present invention. DETAILED DESCRIPTION

[0055] Technical solution content

[0056] A new energy vehicle fault diagnosis and early warning system based on artificial intelligence,

[0057] 1. System Overview

[0058] This system aims to integrate data collection, data analysis, fault diagnosis, warning release, information management and communication modules, and use artificial intelligence algorithms to conduct real-time monitoring and fault warning of key components of new energy vehicles to improve vehicle safety, reliability and maintenance efficiency.

[0059] 2. System Modules and Implementation Methods

[0060] 1. Data acquisition module

[0061] Function: Real-time data collection of key components of new energy vehicles, such as battery voltage (V_bat), battery current (I_bat), motor temperature (T_motor), vehicle speed (V_car) and other parameters;

[0062] Sensor Type:

[0063] Battery voltage sensor: used to collect voltage data (V_bat) of new energy vehicle battery packs in real time. This sensor usually has high precision and stability, and can accurately measure the total voltage of the battery pack and the voltage of each single cell to ensure the normal operation and safety of the battery system;

[0064] Model example: LEM HAH1BVW series

[0065] Such as HAH1BVW S / 12, this sensor has bidirectional current measurement capability and can select different peak current ranges (from ±200A to ±1200A), which is suitable for voltage monitoring of new energy vehicle battery packs.

[0066] Battery current sensor: used to collect current data (I_bat) of new energy vehicle battery packs in real time. The current sensor can detect the charge and discharge current of the battery, ensure the battery's efficiency, and prevent the battery from being overcharged or over-discharged. Hall current sensor is one of the commonly used types. It works based on the Hall effect principle and has the characteristics of isolated measurement. It can realize the isolation conversion of complex signals such as AC, DC, and pulse.

[0067] Model example:

[0068] LEM Hall Current Sensors: Such as LEM’s CAS series or CAB series, these sensors are based on the Hall effect principle and feature high accuracy and isolated measurement, making them suitable for measuring the charge and discharge current of new energy vehicle battery packs.

[0069] Honeywell current sensors: such as the CSNV series, including models CSNV500, CSNV700 and CSNV1500. These sensors have built-in diagnostics and excellent anti-interference performance, which can accurately measure the current of the battery pack and prevent overcharging or over-discharging.

[0070] Motor temperature sensor: used to collect temperature data (T_motor) of new energy vehicle motors in real time. This sensor can monitor the temperature status of the motor to ensure that the motor operates within the normal working range. Once the motor temperature is too high, the system can take timely measures to protect it and prevent damage to the motor;

[0071] Model example:

[0072] NTC temperature sensor: NTC is a semiconductor resistor with a negative temperature coefficient, commonly used to monitor the temperature of motors. The specific model may vary depending on the manufacturer and application scenario, but it usually has high accuracy and stability.

[0073] Pt100 temperature sensor: Pt100 is a platinum metal resistor whose resistance changes linearly with temperature. It is also commonly used for accurate temperature measurement, especially in situations where high-precision measurement is required.

[0074] Vehicle speed sensor: used to collect vehicle speed data (V_car) of new energy vehicles in real time. The vehicle speed sensor is a sensor that measures the speed of electronically controlled vehicles. The control computer uses this input signal to control engine idling, automatic transmission torque converter lock, automatic transmission shifting, engine cooling fan opening and closing, cruise control and other functions;

[0075] Model example:

[0076] The specific model of the vehicle speed sensor may vary depending on the vehicle manufacturer and model. Common vehicle speed sensors include magnetic speed sensors, Hall-type speed sensors, etc.

[0077] For example, some car models may use Hall effect-based vehicle speed sensors that can accurately measure the vehicle's speed and transmit signals to the control computer to achieve various functions such as engine idle speed control, automatic transmission shifting, etc.

[0078] Technical problems solved: Traditional manual detection methods have problems such as long detection cycle, low efficiency, and inaccurate data;

[0079] Technical effect: Through real-time collection, the system can obtain dynamic data during vehicle operation, providing a basis for subsequent data analysis and fault diagnosis;

[0080] Superior effect: It improves the real-time and accuracy of data acquisition and provides strong support for fault warning.

[0081] 2. Data Analysis Module

[0082] Function: preprocessing, feature extraction and anomaly detection of collected data;

[0083] Preprocessing: Use normalization formula to clean and normalize the data to eliminate the dimensional differences between the data;

[0084]

[0085] Feature extraction: extract statistical features such as mean, variance, and peak value to reflect fault characteristics;

[0086] Feature extraction steps

[0087] (1) Determine the characteristic parameters:

[0088] According to the requirements of the fault diagnosis and early warning system for new energy vehicles, the characteristic parameters that need to be extracted are determined. These parameters include the statistical characteristics of key component data such as battery voltage, current, motor temperature, vehicle speed, etc.

[0089] (2) Calculate the mean:

[0090] The mean is calculated for each characteristic parameter. The mean reflects the average level of the data and is an important indicator for evaluating data stability and normality.

[0091] The calculation formula is:

[0092]

[0093] Among them, xi is the i-th data point of the feature parameter, and n is the number of data points.

[0094] (3) Calculate the variance:

[0095] Calculate the variance, which reflects the degree of dispersion of the data, that is, the volatility and stability of the data. A larger variance may mean that the data is abnormal or unstable. The calculation formula is:

[0096]

[0097] (4) Calculate the peak value:

[0098] The peak value reflects the maximum or minimum value in the data and is important for detecting faults and abnormal conditions. For example, a peak value in battery current may mean that there is a fault such as a short circuit or overload.

[0099] The calculation formula is: Peak value = max(x i ) or min(x i )

[0100] (5) Other feature extraction:

[0101] In addition to the mean, variance and peak, other statistical features can be extracted according to actual needs, such as skewness, kurtosis, median, etc. These features can reflect the distribution and characteristics of data from different angles, which helps to more comprehensively evaluate the fault status of new energy vehicles.

[0102] Feature selection and optimization

[0103] After extracting multiple features, feature selection and optimization may be required. The purpose of this step is to screen out the most sensitive and effective features for fault diagnosis and early warning, so as to reduce computational complexity and improve diagnostic accuracy. Feature selection and optimization can be achieved through various methods, such as feature selection based on correlation analysis, feature selection based on machine learning algorithms, etc.

[0104] Through the above steps, the feature extraction work in the data analysis module can be completed. These extracted features will serve as important inputs for subsequent fault diagnosis and early warning models, providing strong guarantees for the safe operation of new energy vehicles.

[0105] Anomaly detection: Detect outliers in data using the anomaly index (AI) calculation formula;

[0106]

[0107] Technical problems solved: Noise, outliers and other issues exist in the original data, which affect the accuracy of subsequent fault diagnosis;

[0108] Technical effect: The quality and availability of data are improved through preprocessing and feature extraction; abnormal values ​​in data can be found in time through anomaly detection, providing clues for subsequent fault diagnosis;

[0109] Superior effect: Improved the accuracy and efficiency of data analysis, and provided reliable data support for fault diagnosis.

[0110] 3. Fault diagnosis module

[0111] Function: Use machine learning models to diagnose abnormal data and output the probability distribution of fault type and fault degree;

[0112] Model structure: includes input layer, hidden layer and output layer; the input layer receives abnormal data; the hidden layer uses RELU activation function for nonlinear transformation; the output layer uses softmax function to output the probability distribution of fault type and fault degree;

[0113] RELU(x)=max(0,x)

[0114] The output layer uses the softmax function to output the probability distribution of fault type and fault degree. The softmax function is defined as:

[0115]

[0116] Technical problems solved: Traditional fault diagnosis methods rely on expert experience and manual testing, which have problems such as long diagnosis cycle and low accuracy;

[0117] Technical effect: Through the machine learning model, automatic identification of fault types and quantitative evaluation of fault severity are achieved;

[0118] Superior effect: Improved the accuracy and efficiency of fault diagnosis and reduced dependence on expert experience.

[0119] 4. Early warning module

[0120] Function: Receive the diagnosis results of the fault diagnosis module, formulate corresponding early warning strategies according to the fault type and severity, and generate specific early warning information;

[0121] Specifically, they include:

[0122] Early warning strategy formulation unit: This unit is responsible for intelligently formulating and optimizing corresponding early warning strategies based on the type of fault diagnosed and its severity; specific strategies include but are not limited to using sound and light alarms (such as flashing warning lights and buzzer alarms on the vehicle dashboard), receiving SMS notifications through the owner's mobile phone, and pushing early warning information on the owner's smartphone APP. These strategies are formulated to ensure that the owner can quickly and accurately obtain fault information so that he can take appropriate countermeasures.

[0123] Warning information generation unit: This unit is responsible for generating specific and detailed warning information based on the strategy formulated by the warning strategy formulation unit; the content of the warning information includes but is not limited to: clearly stated fault type (such as abnormal battery voltage, motor temperature is too high, etc.), detailed description of the fault degree (such as minor, moderate or severe), and recommended response measures for the fault (such as immediate parking inspection, contacting a maintenance station, etc.). The generation of these warning information is intended to help car owners quickly understand the fault situation and make correct response decisions.

[0124] Technical problems to be solved: Traditional early warning methods have problems such as untimely warnings and inaccurate warning information;

[0125] Technical effect: Through the formulation of early warning strategies and the generation of early warning information, timely early warning and accurate prompts of faults are achieved;

[0126] Superior effect: Improves the timeliness and accuracy of early warning, helping drivers or maintenance personnel to take timely measures to avoid the expansion of faults.

[0127] 5. Information management module

[0128] Function: Store, manage and display vehicle operation data, fault history and warning information;

[0129] Specifically, they include:

[0130] Data storage unit: This unit is mainly responsible for storing various data generated during the operation of new energy vehicles, including but not limited to vehicle operation data (such as battery voltage, current, motor temperature, vehicle speed and other real-time data), fault history records (detailed information such as the results, time, location, etc. of each fault diagnosis), and warning information (including warning time, warning type, fault description and recommended measures, etc.). To ensure efficient management and fast access to data, these data are stored in a structured database, such as a relational database or a non-relational database. The data are associated through reasonable fields and indexes to facilitate subsequent query, analysis and statistics.

[0131] Data display unit: This unit is responsible for displaying the data stored in the data storage unit to the driver or maintenance personnel in an intuitive and easy-to-understand form. The display forms include but are not limited to various charts (such as line charts, bar charts, pie charts, etc., used to display the changing trends of vehicle performance parameters, the statistics of fault distribution, etc.) and reports (such as daily reports, weekly reports, monthly reports, etc., used to summarize and display the operating status and fault conditions of the vehicle over a period of time). Through these charts and reports, drivers and maintenance personnel can clearly understand the current status of the vehicle and the past fault history, so as to make more accurate judgments and decisions. At the same time, the data display unit also has a data export function, which can export the displayed data in Excel, PDF and other formats to facilitate further analysis and archiving.

[0132] Technical problems solved: Traditional information management methods have problems such as difficulty in data query and non-intuitive information display;

[0133] Technical effect: Through data storage and display, comprehensive management and visual display of vehicle operation data are realized;

[0134] Superior effect: Improves the convenience and intuitiveness of information management, helping drivers or maintenance personnel to better understand the vehicle status.

[0135] 6. Communication module

[0136] Function: Realize data communication between modules and data transmission between the system and external devices;

[0137] Specifically,

[0138] The communication module uses advanced wireless communication technology, including but not limited to 4G / 5G mobile communication networks, Wi-Fi wireless LAN and Bluetooth short-range communication technology, to achieve efficient and stable data communication between modules in the new energy vehicle fault diagnosis and early warning system, and ensure data transmission between the system and external devices (such as the owner's smartphone, the server of the maintenance station, etc.). During the data transmission process, the communication module uses encryption protocols to encrypt the data to ensure the security and privacy protection of the data during the transmission process. These encryption protocols include but are not limited to SSL / TLS protocols, AES encryption algorithms, etc., which can effectively prevent data from being stolen, tampered with or leaked during the transmission process, thereby protecting the owner's personal privacy and the security of vehicle data.

[0139] In addition, the communication module also has an adaptive network switching function, which can intelligently select the optimal communication method according to the current network environment and communication requirements to improve the efficiency and stability of data transmission. At the same time, the communication module also supports breakpoint resumption and data verification functions to ensure data integrity and reliability.

[0140] Technical problems solved: Traditional communication methods have problems such as slow data transmission speed and poor data security;

[0141] Technical effect: Through wireless communication technology (such as 4G / 5G, Wi-Fi or Bluetooth) and encryption protocols, high-speed and secure data transmission is achieved;

[0142] Superior effect: Improved the efficiency and security of data transmission and ensured the stable operation of the system.

[0143] Conclusion

[0144] This system integrates multiple modules and uses artificial intelligence algorithms to perform real-time monitoring and fault warning of new energy vehicles. The modules work together in function to solve the problems existing in traditional fault diagnosis and warning methods. By implementing this system, the safety, reliability and maintenance efficiency of new energy vehicles can be significantly improved, providing strong support for the popularization and development of new energy vehicles.

[0145] Specific application example 1:

[0146] This system mainly includes data acquisition module, data analysis module, fault diagnosis module, early warning module, information management module and communication module.

[0147] Specific application examples

[0148] 1. Data acquisition module

[0149] Function: Real-time collection of key component data of new energy vehicles such as battery voltage (V_bat), battery current (I_bat), motor temperature (T_motor), and vehicle speed (V_car).

[0150] Data collection method

[0151] The data acquisition module usually collects data in the following ways:

[0152] Sensor connection: Each key component of new energy vehicles, such as batteries and motors, is equipped with corresponding sensors, which can monitor the working status of the components in real time and convert relevant data into electrical signals. The data acquisition module connects to these sensors, receives and processes these electrical signals, and thus obtains real-time data of key components;

[0153] CAN bus communication: Modern new energy vehicles generally use CAN (Controller Area Network) bus technology to achieve communication between various controllers inside the vehicle. By connecting to the CAN bus, the data acquisition module can obtain data sent by various controllers of the vehicle in real time, including battery voltage, battery current, motor temperature, vehicle speed, etc.

[0154] Data collection process

[0155] The specific process of data collection by the data collection module is as follows:

[0156] Signal reception: The data acquisition module first receives signals from sensors or CAN bus. These signals are usually analog or digital signals, representing real-time data of key components;

[0157] Signal processing: After receiving the signal, the data acquisition module will process the signal. This includes steps such as signal amplification, filtering, analog-to-digital conversion, etc. to ensure the accuracy and stability of the signal;

[0158] Data storage and transmission: The processed data will be stored by the data acquisition module and transmitted to the central processor or cloud platform through a specific communication protocol (such as CAN protocol, Ethernet protocol, etc.) for further analysis and processing;

[0159] The data acquisition module collects data of key components of new energy vehicles in real time by connecting sensors and CAN bus, and provides strong data support for subsequent fault diagnosis and early warning through signal processing, storage and transmission.

[0160] Example: When a new energy vehicle is driving, the data acquisition module collects in real time the battery voltage of 350V, the battery current of 100A, the motor temperature of 80℃, and the vehicle speed of 60km / h.

[0161] 2. Data Analysis Module

[0162] Function: preprocessing, feature extraction and anomaly detection of collected data;

[0163] Preprocessing: Use normalization formula to clean and normalize the data;

[0164]

[0165] Among them, x i is the original data, x min and x max are the minimum and maximum values ​​in the data set, respectively, and x norm，i is the normalized data;

[0166] Feature extraction: extract statistical features such as mean, variance, and peak value to reflect fault characteristics;

[0167] Anomaly detection: Detect outliers in data using the anomaly index (AI) calculation formula;

[0168]

[0169] Among them, xi is the value of the ith data point, μ i is the historical average value of the data point, σ i is the historical standard deviation of the data point, and n is the number of data points; when AI exceeds the preset threshold, it is judged as abnormal;

[0170] Example: After the data analysis module preprocesses the collected data, the normalized battery voltage is 0.7 (assuming the minimum value in the data set is 0V and the maximum value is 500V), the battery current is 0.5 (assuming the minimum value in the data set is 0A and the maximum value is 200A), and the motor temperature is 0.6 (assuming the minimum value in the data set is 0℃ and the maximum value is 120℃). Then, the mean, variance and other features are extracted, and the AI ​​value is calculated using the abnormal index formula to be 1.5, which exceeds the preset threshold of 1 and is determined to be abnormal data.

[0171] 3. Fault diagnosis module

[0172] Function: Use machine learning models to diagnose abnormal data and output the probability distribution of fault type and fault degree;

[0173] Model structure: includes input layer, hidden layer and output layer; the input layer receives abnormal data; the hidden layer uses RELU activation function for nonlinear transformation; the output layer uses softmax function to output the probability distribution of fault type and fault degree.

[0174] RELU(x)=max(0,x)

[0175]

[0176] Among them, z j is the output of the jth neuron, K is the total number of neurons in the output layer, that is, the number of fault types;

[0177] Example: After receiving abnormal data, the fault diagnosis module uses the machine learning model to perform fault diagnosis, and the output result is a battery overheating fault with a probability of 90%.

[0178] 4. Early warning module

[0179] Function: Receive the diagnosis results of the fault diagnosis module, formulate corresponding early warning strategies according to the fault type and severity, and generate specific early warning information;

[0180] Example: After receiving the output results of the fault diagnosis module, the early warning module formulates an early warning strategy to stop the vehicle immediately to check the battery and generates an early warning message: "The battery is overheated, please stop the vehicle immediately to check!" The driver is warned through sound and light alarms, text messages, or APP push.

[0181] 5. Information management module

[0182] Function: Store, manage and display vehicle operation data, fault history and warning information;

[0183] Example: The information management module stores the collected data, fault diagnosis results and warning information in the database, and provides query and display functions. Drivers or maintenance personnel can use the query function to view the vehicle's operating status, historical fault records and warning information.

[0184] 6. Communication module

[0185] Function: Realize data communication between modules and data transmission between the system and external devices;

[0186] Example: The communication module uses wireless communication technology (such as 4G / 5G, Wi-Fi or Bluetooth) to realize data communication between modules and transmit warning information to devices such as the driver's mobile phone or vehicle display.

[0187] System Effects

[0188] Through the application of this system, real-time monitoring, fault diagnosis and early warning of key components of new energy vehicles can be realized; when a vehicle fails, the system can promptly detect and issue early warning information, reminding the driver or maintenance personnel to take appropriate measures to deal with it, which can not only improve the safety and reliability of the vehicle, but also reduce maintenance costs and improve maintenance efficiency; at the same time, this system can also provide data support and optimization suggestions for the research and development and production of new energy vehicles, and promote the sustainable development of the new energy vehicle industry.

[0189] Case 1: Battery system fault warning and diagnosis

[0190] background:

[0191] While a new energy vehicle was driving, the owner suddenly found a warning message of battery system failure on the display screen and the vehicle's power decreased.

[0192] System application refinement:

[0193] Data collection:

[0194] The system collects key parameters of the battery such as voltage (V_bat), current (I_bat), and temperature (T_bat) in real time.

[0195] Assume that the voltage collected at a certain moment is 300V, the current is 100A, and the temperature is 40°C (these values ​​are only examples, and the actual values ​​will vary depending on the vehicle and battery status).

[0196] Data Analysis:

[0197] Use anomaly detection algorithms, such as the statistically based anomaly index (AI) calculation formula:

[0198]

[0199] Among them, x i is the value of the ith data point, μ i is the historical average value of the data point, σ i is the historical standard deviation of the data point, and n is the number of data points; when AI exceeds the preset threshold, it is judged as abnormal.

[0200] Assume that the historical average voltage is 320V and the historical standard deviation is 10V; the historical average current is 120A and the historical standard deviation is 20A; the historical average temperature is 35℃ and the historical standard deviation is 5℃.

[0201] For voltage, we calculate:

[0202]

[0203] For current and temperature, the same calculation is performed to obtain the AI ​​value.

[0204] If the AI ​​value exceeds the preset threshold (such as 1.5), it is judged as abnormal.

[0205] Troubleshooting:

[0206] Based on the AI ​​value, the system determines that the battery voltage is abnormally low and there may be a fault.

[0207] Further diagnosis using a machine learning model determined that the performance of one of the cells in the battery pack had severely degraded.

[0208] Warning issued:

[0209] The system sends a warning message to the car owner, indicating a battery system failure and recommends immediate parking and inspection.

[0210] Processing results:

[0211] The owner sent the vehicle to the repair station, where technicians replaced the faulty single battery based on the system's diagnosis results and balanced the battery pack. After the repairs were completed, the vehicle resumed normal driving.

[0212] Case 2: Motor drive system fault warning and diagnosis

[0213] background:

[0214] When a new energy vehicle was accelerating, the owner found that the acceleration was unstable and a warning message of electric drive system failure appeared on the display screen.

[0215] System application refinement:

[0216] Data collection:

[0217] The system collects the voltage (V_mcu), current (I_mcu) and other parameters of the motor control unit in real time.

[0218] Assume that the voltage collected at a certain moment is 200V and the current is 250A (these values ​​are only examples).

[0219] Data Analysis:

[0220] Using feature extraction and anomaly detection algorithms, such as sliding window-based anomaly detection, it is found that the current value fluctuates greatly during the acceleration process.

[0221] Calculate the standard deviation or coefficient of variation of the current to determine its stability.

[0222] Troubleshooting:

[0223] Using machine learning models or rule-based expert systems, a poor contact of a circuit connection plug in the motor control unit was diagnosed.

[0224] Warning issued:

[0225] The system sends a warning message to the car owner, indicating that the electric drive system is faulty and recommends repairs as soon as possible.

[0226] Processing results:

[0227] The owner took the vehicle to the repair station, where technicians reconnected the plug of the faulty circuit and performed a functional test.

[0228] The above cases demonstrate the effect and value of the artificial intelligence-based new energy vehicle fault diagnosis and early warning system in practical applications. Through real-time monitoring and diagnosis, the system can promptly detect and warn of potential faults, improve vehicle safety and reliability, reduce maintenance costs and improve maintenance efficiency.

[0229] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A new energy vehicle fault diagnosis and early warning system based on artificial intelligence, characterized in that: include: a) Data acquisition module, which collects driving data of key components and characteristic indicators of driving conditions of new energy vehicles in real time; b) A data analysis module receives data transmitted by the data acquisition module and analyzes the data using an artificial intelligence algorithm; c) Fault diagnosis module, which uses machine learning models to perform fault diagnosis based on the results of the data analysis module and outputs the fault type and fault severity; d) Early warning module, which receives the diagnosis result of the fault diagnosis module and sends early warning information to the driver or maintenance personnel when a serious or urgent fault is detected; e) Information management module, used to store, manage and display vehicle operation data, fault history records and warning information; f) Communication module, which realizes data communication between modules and data transmission between the system and external devices.

2. The artificial intelligence-based new energy vehicle fault diagnosis and early warning system according to claim 1, characterized in that: The data acquisition module includes multiple sensors, specifically: Battery voltage sensor: collects voltage data of new energy vehicle battery packs in real time; Battery current sensor: collects current data of new energy vehicle battery packs in real time; Motor temperature sensor: collects temperature data of new energy vehicle motors in real time; Vehicle speed sensor: collects vehicle speed data of new energy vehicles in real time.

3. The artificial intelligence-based new energy vehicle fault diagnosis and early warning system according to claim 1 is characterized in that: The data analysis module uses the following formula to calculate the abnormal index of the data: Among them, x i is the value of the ith data point, μ i is the historical average value of the data point, σ i is the historical standard deviation of the data point, and n is the number of data points; when AI exceeds the preset threshold, it is judged as abnormal.

4. The artificial intelligence-based new energy vehicle fault diagnosis and early warning system according to claim 3 is characterized in that: The data analysis module also includes: The data preprocessing unit is used to clean, denoise and normalize the collected data. The following formula is used for normalization: Among them, x i is the original data, x min and x max are the minimum and maximum values ​​in the data set, respectively, and x norm，i is the normalized data; A feature extraction unit extracts key information that can reflect fault characteristics from the preprocessed data; The anomaly detection unit uses the above anomaly index calculation formula to perform anomaly detection on the data after feature extraction, and outputs the abnormal data to the fault diagnosis module.

5. The artificial intelligence-based new energy vehicle fault diagnosis and early warning system according to claim 1 is characterized in that: The machine learning model of the fault diagnosis module further includes: Input layer, receives abnormal data, and the number of neurons in the input layer is consistent with the number of features extracted by the feature extraction unit; The hidden layer contains multiple layers of neurons. The number of neurons in each layer is determined according to the model complexity and the size of the data set. The neurons in the hidden layer use the ReLU activation function for nonlinear transformation. The RELU function is defined as: RELU(x)=max(0,x) The output layer uses the softmax function to output the probability distribution of fault type and fault degree. The softmax function is defined as: Among them, z j is the output of the jth neuron, and K is the total number of neurons in the output layer, that is, the number of fault types.

6. The artificial intelligence-based new energy vehicle fault diagnosis and early warning system according to claim 1 is characterized in that: The early warning module also includes: The early warning strategy formulation unit formulates corresponding early warning strategies according to the fault type and severity, including the use of sound and light alarms, SMS notifications or APP push; The warning information generation unit generates specific warning information according to the warning strategy, including the fault type, fault degree and recommended measures.

7. The artificial intelligence-based new energy vehicle fault diagnosis and early warning system according to claim 1 is characterized in that: The information management module also includes: Data storage unit, used to store vehicle operation data, fault history records and warning information. The data is stored in a structured database for easy query and analysis; The data display unit displays the stored data to the driver or maintenance personnel in the form of charts and reports, providing intuitive vehicle status monitoring and fault analysis.

8. The artificial intelligence-based new energy vehicle fault diagnosis and early warning system according to claim 1 is characterized in that: The communication module adopts wireless communication technology to realize data communication between modules and data transmission between the system and external devices. The data transmission adopts encryption protocol to ensure data security and privacy protection.

Images