New energy automobile intelligent network connection remote diagnosis and fault early warning system

Abstract

The invention discloses a new energy automobile intelligent network connection remote diagnosis and fault early warning system, and relates to the technical field of new energy automobiles, the system comprises a data acquisition module, a data transmission module, a data analysis and processing module, a personalized early warning model construction module and a fault diagnosis and early warning module; the data acquisition module is used for acquiring driving habit data, vehicle use environment data and historical fault record data of the new energy vehicle; according to the method, a personalized fault early warning model is established for each vehicle by collecting and analyzing multi-dimensional data such as user driving habits, vehicle use environments and historical fault records, accurate early warning of new energy vehicle faults is realized, and by collecting and analyzing the multi-dimensional data, the personalized difference of vehicle use is fully considered, so that the accuracy of new energy vehicle fault early warning is improved. Compared with a traditional general fault early warning system, the accuracy and timeliness of fault early warning are greatly improved, and the situations of false alarm and missing alarm are effectively avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy vehicles, and in particular to an intelligent network-connected remote diagnosis and fault warning system for new energy vehicles. Background Art

[0002] New energy vehicles refer to new types of vehicles that use unconventional automotive fuels as their power source (such as electricity, hydrogen, etc.) and integrate advanced vehicle power control and drive technologies. They have the advantages of environmental protection, high efficiency, and energy diversification, and are gradually becoming the mainstream direction of development in the automotive industry. Diagnosis and fault warning of new energy vehicles are of great significance. The core systems of new energy vehicles (such as battery management systems, motor control systems, etc.) are complex in structure and have significant differences in fault types and manifestations from traditional fuel vehicles. For example, if problems such as thermal runaway and charging failures of the battery system are not discovered and handled in a timely manner, it may lead to a decline in vehicle performance and even cause safety accidents. In addition, new energy vehicles are highly intelligent, and their electronic control systems involve a large number of precision components. Failure in any link may affect the overall operation of the vehicle. Therefore, timely and accurate fault diagnosis and warning are crucial to ensuring the safe and reliable operation of new energy vehicles and improving user experience.

[0003] According to patent application number 202510039324.4, an artificial intelligence-based fault diagnosis and early warning system for new energy vehicles is disclosed, including: a data acquisition module, which collects driving data of characteristic indicators of key components and driving conditions of new energy vehicles in real time; a data analysis module, which receives data transmitted by the data acquisition module and analyzes the data using artificial intelligence algorithms; a fault diagnosis module, which uses a machine learning model to diagnose faults based on the results of the data analysis module and outputs the fault type and fault severity. The above solution can significantly improve the safety, reliability and maintenance efficiency of new energy vehicles, reduce the maintenance costs of car owners, and improve the user experience.

[0004] However, the existing new energy vehicle fault warning systems still have certain defects when used. Most of the existing new energy vehicle fault warning systems use general fault diagnosis rules and thresholds, and do not fully consider personalized factors such as the driving habits of different users, the actual vehicle usage environment, and historical fault records. This leads to false alarms and missed alarms in actual applications. For example, vehicles that frequently travel short distances in cold areas have a battery degradation rate that is different from that in normal usage scenarios. General fault warning rules cannot accurately judge the battery health status. Users with different driving styles have different losses in the vehicle power system and braking system. The existing system is difficult to perform targeted fault warnings, which seriously affects user experience and driving safety. Therefore, it is of great significance to develop a new energy vehicle intelligent network remote diagnosis and fault warning system. Summary of the Invention

[0005] The purpose of the present invention is to make up for the shortcomings of the existing technology and provide a new energy vehicle intelligent network remote diagnosis and fault warning system, which can achieve accurate warning of new energy vehicle faults by establishing a personalized warning model based on user driving habits, vehicle usage environment and historical fault records. By collecting and analyzing multi-dimensional data, it fully considers the personalized differences in vehicle usage, improves the accuracy and timeliness of fault warnings, and effectively avoids the occurrence of false alarms and missed alarms. It uses intelligent network technology to realize real-time transmission and remote diagnosis of data, so that car owners can timely understand the potential fault information of the vehicle and take corresponding measures in advance to ensure driving safety. Vehicle manufacturers can also optimize product design and after-sales service according to the fault diagnosis results, thereby improving user satisfaction and corporate competitiveness.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a new energy vehicle intelligent network remote diagnosis and fault warning system, the system includes: a data acquisition module, a data transmission module, a data analysis and processing module, a personalized warning model construction module and a fault diagnosis and warning module;

[0007] The data acquisition module is used to collect driving habit data, vehicle usage environment data and historical fault record data of new energy vehicles;

[0008] The data transmission module is used to transmit the data acquired by the data acquisition module to the cloud server in real time;

[0009] The data analysis and processing module is deployed on the cloud server and is used to analyze the transmitted data to obtain a driving habit feature model, a vehicle usage environment feature model, and a historical fault correlation model;

[0010] The personalized warning model building module establishes a personalized fault warning model for each vehicle based on the driving habit feature model, the vehicle usage environment feature model, and the historical fault association model;

[0011] The fault diagnosis and warning module monitors the operating status data of various vehicle systems and components in real time, inputs it into the corresponding personalized warning model for analysis, and when the warning threshold is reached or exceeded, sends fault warning information to the owner's mobile terminal through the data transmission module, and feeds back the fault diagnosis results to the vehicle manufacturer's after-sales service center.

[0012] Furthermore, in the data acquisition module, driving habit data is collected through displacement sensors and pressure sensors installed on the accelerator pedal, brake pedal, and steering wheel, including the accelerator pedal travel change rate, brake pedal stepping force and frequency, and steering angle change rate. Vehicle usage environment data is collected by environmental sensors and GPS positioning modules. The environmental sensors include temperature sensors, humidity sensors, and air pressure sensors. The collected data includes ambient temperature, humidity, and altitude. The GPS positioning module obtains the vehicle's driving area and driving time.

[0013] Furthermore, the historical fault record data is recorded by the vehicle electronic control unit and stored in a local database. The data acquisition module reads it in real time and performs preliminary preprocessing of filtering and noise reduction after acquisition. The preprocessing adopts a dynamic weighted filtering algorithm to dynamically adjust the weight to achieve adaptive filtering of abnormal data. The calculation formula is: ,in, For the Second filter output value, For the The original data collected, is the dynamic weight coefficient at the current moment, satisfying is the current data fluctuation range, is the preset fluctuation threshold, To adjust the parameters, is the historical data weight coefficient, is the historical data window size.

[0014] Furthermore, the data transmission module adopts 4G / 5G technology to encapsulate the driving habit data, vehicle usage environment data and historical fault record data processed by the data acquisition module into data packets according to the TCP / IP communication protocol and transmit them to the cloud server.

[0015] Furthermore, the data analysis and processing module uses a clustering algorithm to perform cluster analysis on the driving habit data, classifying driving styles into aggressive, stable, and economical types, and calculates corresponding driving habit characteristic parameters to establish a driving habit characteristic model. An association rule algorithm is used to analyze the vehicle usage environment data to establish a vehicle usage environment characteristic model. A time series analysis algorithm is used to analyze historical fault record data, and a historical fault association model is established by combining the driving habits and usage environment data. When establishing the driving habit characteristic model, a driving style quantitative evaluation formula is used to quantitatively score driving behavior to classify driving style types. The formula is: ,in, Quantify the driving style score, for The positive rate of change of the accelerator pedal at any moment, for The brake pedal is pressed hard at all times. for The change in steering angle at any moment, is the data collection period, 、 、 is the weight coefficient, and .

[0016] Furthermore, the personalized warning model construction module is based on a deep learning framework and adopts a long short-term memory network to construct a model. It takes the data of the driving habit feature model, the vehicle usage environment feature model and the historical fault association model as input, and uses the actual fault data of various vehicle systems and components as training labels for training. By adjusting the learning rate and the number of hidden layer neurons parameters, the model prediction error is minimized to obtain a personalized fault warning model.

[0017] Furthermore, when training the model, a dynamic prediction formula for fault probability is used to achieve dynamic prediction of fault probability through multi-feature fusion. The formula is: ,in, for Always based on historical fault data , environmental data , driving habit data The probability of failure, is the feature weight parameter, is a joint feature function that integrates historical fault correlation, environmental influencing factors and driving habit characteristics. 、 、 For the Feature vector of the class fault.

[0018] Furthermore, the fault diagnosis and warning module obtains the operating status data of various vehicle systems and components from the vehicle electronic control unit in real time, inputs the data into the personalized warning model, and when the model predicts that the probability of a fault exceeds a set threshold, it generates fault warning information including the fault type, possible cause, and recommended measures, which is sent to the owner's mobile terminal through the data transmission module, and the fault diagnosis results are uploaded to the vehicle manufacturer's after-sales service center database.

[0019] Furthermore, after the data acquisition module collects data, it performs preliminary pre-processing of filtering and noise reduction on the data, and then transmits it to the cloud server by the data transmission module. The fault warning information sent by the fault diagnosis and warning module is sent to the car owner's mobile terminal in the form of text messages and push notifications.

[0020] Compared with existing technologies, this new energy vehicle intelligent network remote diagnosis and fault warning system has the following beneficial effects:

[0021] 1. The present invention establishes a personalized fault warning model for each vehicle by collecting and analyzing multi-dimensional data such as user driving habits, vehicle usage environment, and historical fault records, thereby achieving accurate early warning of new energy vehicle faults. By collecting and analyzing multi-dimensional data, the personalized differences in vehicle usage are fully considered. Compared with traditional general fault warning systems, the accuracy and timeliness of fault warnings are greatly improved, and the occurrence of false alarms and missed alarms is effectively avoided.

[0022] 2. The present invention uses intelligent network technology to achieve real-time data transmission and remote diagnosis. Vehicle owners can promptly understand potential vehicle fault information and take corresponding measures in advance to ensure driving safety. Vehicle manufacturers can also optimize product design and after-sales service based on fault diagnosis results, improve user satisfaction and corporate competitiveness, and use advanced algorithms such as deep learning to build personalized early warning models that can dynamically adapt to various changes in the vehicle operation process. With the continuous accumulation and learning of data, the model's early warning capability will continue to improve, with good adaptability and scalability.

[0023] Other advantages, objects and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be learned from the practice of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0025] Figure 1 This is a structural diagram of a new energy vehicle intelligent network remote diagnosis and fault warning system;

[0026] Figure 2 This is a workflow diagram of a new energy vehicle intelligent network remote diagnosis and fault warning system. DETAILED DESCRIPTION

[0027] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0028] See also Figure 1A new energy vehicle intelligent network remote diagnosis and fault warning system consists of a data acquisition module, a data transmission module, a data analysis and processing module, a personalized warning model construction module and a fault diagnosis and warning module. The data acquisition module collects driving habit data, vehicle usage environment data and historical fault record data of new energy vehicles in real time. Through the displacement sensors and pressure sensors installed on the accelerator pedal, brake pedal and steering wheel, the accelerator pedal travel change rate, brake pedal stepping force and frequency, steering angle change rate and other data are obtained. The environmental sensors (temperature, humidity, air pressure sensors) are used to collect the ambient temperature, humidity and altitude. The GPS positioning module is combined to obtain the vehicle driving area and time. The data is read from the local database of the vehicle electronic control unit in real time. After collection, the data is filtered and noise reduction preprocessed.

[0029] The data transmission module uses 4G / 5G technology to encapsulate the driving habits, environment, and historical fault data processed by the data acquisition module into data packets according to the TCP / IP communication protocol and transmit them to the cloud server in real time.

[0030] The data analysis and processing module uses a clustering algorithm to analyze driving data, classifying driving styles into aggressive, stable, and economical. It then establishes a feature model based on a quantitative evaluation system, uses an association rule algorithm to analyze the relationship between environmental data (temperature, humidity, etc.) and vehicle operation, and constructs an environmental impact model. Using a time series analysis algorithm, combined with driving habits and environmental data, it explores the patterns and associated factors of historical faults.

[0031] The personalized warning model construction module is based on a deep learning framework and adopts a long short-term memory network. It takes driving habits, environmental characteristics, and historical fault association model data as input, and actual fault data as training labels. By adjusting the parameters and optimizing the model, a personalized fault warning model is established for each vehicle.

[0032] The fault diagnosis and warning module obtains the operating data of various vehicle systems and components in real time, inputs it into a personalized model for analysis, and generates warning information containing the fault type, cause, and recommended measures when the fault probability exceeds the threshold. The information is sent to the owner's mobile terminal via SMS and push notification, and the diagnostic results are fed back to the vehicle manufacturer's after-sales service center.

[0033] Example 1

[0034] This embodiment is applied to a pure electric new energy vehicle equipped with the new energy vehicle intelligent network remote diagnosis and fault warning system. It is mainly used to provide accurate warnings of possible vehicle faults during daily driving to ensure driving safety. At the same time, it provides fault diagnosis data to vehicle manufacturers to optimize product design and after-sales service. Figure 2 The specific implementation process is as follows:

[0035] In this pure electric new energy vehicle, the data acquisition module starts working, and the displacement sensors and pressure sensors installed on the accelerator pedal, brake pedal, and steering wheel continuously collect driving habit data. Among them, the accelerator pedal stroke change rate is obtained by real-time monitoring of the accelerator pedal position changes at different times through the displacement sensor. The brake pedal stepping force and frequency are sensed by the pressure sensor to sense the force of each step and the number of steppings per unit time. The steering angle change rate is recorded by the sensor at the steering wheel to record the change of the steering angle in the time dimension.

[0036] The vehicle's environmental data is collected by environmental sensors and the GPS positioning module. The temperature sensor, humidity sensor, and air pressure sensor in the environmental sensor respectively detect the ambient temperature, humidity, and air pressure around the vehicle in real time, thereby obtaining data such as ambient temperature, humidity, and altitude; the GPS positioning module obtains the vehicle's driving area and time information in real time.

[0037] The historical fault record data is stored in the local database of the vehicle electronic control unit. The data acquisition module will read this data in real time. After collecting all the data, the data acquisition module will perform preliminary preprocessing of filtering and noise reduction on the data. Here, the dynamic weighted filtering algorithm is used, and its calculation formula is: ,in, For the Second filter output value, For the The original data collected, is the dynamic weight coefficient at the current moment, which satisfies is the current data fluctuation range, is the preset fluctuation threshold, To adjust the parameters, is the historical data weight coefficient, is the size of the historical data window. Through this algorithm, the weight is dynamically adjusted according to the fluctuation of the data to achieve adaptive filtering of abnormal data and improve the accuracy of the data.

[0038] The processed data is encapsulated into data packets by the data transmission module using 4G technology according to the TCP / IP communication protocol and transmitted to the cloud server in real time. After receiving the data, the data analysis and processing module deployed on the cloud server begins to analyze the data.

[0039] For driving habit data, cluster analysis is performed using clustering algorithms. First, the driving style quantitative evaluation formula is used. Quantitative scoring of driving behavior, including: Quantify the driving style score, for The positive rate of change of the accelerator pedal at any moment, for The brake pedal is pressed hard at all times. for The change in steering angle at any moment, is the data collection period, 、 、 is the weight coefficient, and ,Through the calculated quantitative scores, the driving styles are divided into aggressive, stable and economical types, and the corresponding driving habit characteristic parameters are calculated to establish a driving habit characteristic model.

[0040] For vehicle usage environment data, association rule algorithms are used to analyze the correlation between environmental data and vehicle operating status, and a vehicle usage environment feature model is established.

[0041] For historical fault record data, we use time series analysis algorithms to analyze it, and combine it with driving habits and usage environment data to explore the patterns of historical fault occurrence and their correlation with other factors, and establish a historical fault correlation model.

[0042] The personalized warning model construction module is based on a deep learning framework and uses a long short-term memory network (LSTM) to build the model. It takes data from the driving habit feature model, the vehicle usage environment feature model, and the historical fault association model as input, and uses the actual fault data of various vehicle systems and components as training labels for training. During the training process, the model's prediction error is minimized by adjusting parameters such as the learning rate and the number of hidden layer neurons, thereby establishing a personalized fault warning model for the vehicle.

[0043] When training the model, the dynamic prediction formula of fault probability is used To achieve dynamic prediction of failure probability, for Always based on historical fault data , environmental data , driving habit data The probability of failure, is the feature weight parameter, is a joint feature function that integrates historical fault correlation, environmental influencing factors and driving habit characteristics. 、 、 For the The characteristic vector of the class fault is obtained by fusing multiple feature information through this formula to achieve dynamic prediction of fault probability.

[0044] The fault diagnosis and warning module obtains the operating status data of various vehicle systems and components from the vehicle's electronic control unit in real time, and inputs this data into the corresponding personalized warning model for analysis. When the fault probability predicted by the model reaches or exceeds the warning threshold, the fault diagnosis and warning module will generate fault warning information including the fault type, possible cause, and recommended measures. These warning information are sent to the owner's mobile terminal in the form of SMS and push notifications through the data transmission module, and the fault diagnosis results are fed back to the vehicle manufacturer's after-sales service center.

[0045] To sum up, through the specific implementation process of this embodiment, the new energy vehicle intelligent network remote diagnosis and fault warning system fully utilizes the functions of each module. The entire system fully considers the personalized differences in vehicle use, greatly improves the accuracy and timeliness of fault warnings, and effectively avoids the occurrence of false alarms and missed alarms. Car owners can timely understand the potential fault information of the vehicle and take corresponding measures in advance to ensure driving safety; vehicle manufacturers can also optimize product design and after-sales service based on the fault diagnosis results, thereby improving user satisfaction and corporate competitiveness. With the continuous accumulation of vehicle operation data, the warning capability of the system will continue to improve, and it has good adaptability and scalability.

[0046] Example 2

[0047] This embodiment is applied to a commercial logistics vehicle fleet of a certain new energy vehicle brand. The fleet consists of multiple vehicles equipped with the same intelligent network remote diagnosis and fault warning system. It is mainly used to provide real-time fault warnings for key components such as the battery system and motor control system of the fleet vehicles in logistics and transportation scenarios, optimize fleet operating efficiency, reduce maintenance costs, and provide vehicle manufacturers with batch vehicle fault data to improve the reliability design of commercial vehicle models. Figure 2 The specific implementation process is as follows:

[0048] In this fleet of new energy commercial logistics vehicles, the data collection modules of each vehicle are deployed according to unified standards. In terms of driving habit data collection, the displacement sensor at the accelerator pedal continuously monitors the accelerator pedal stroke change rate of the logistics vehicle during cargo transportation. Since logistics vehicles are often overloaded, their stroke change characteristics are different from those of passenger cars; the pressure sensor at the brake pedal synchronously records the pedaling force under different road conditions (such as frequent braking on urban roads and light braking on highways) and frequency, and the sensor at the steering wheel tracks the steering angle change rate of the logistics vehicle when turning and changing lanes, especially the steering operation characteristics under heavy load conditions.

[0049] When collecting vehicle usage environment data, the temperature sensor in the environmental sensor focuses on monitoring the real-time temperature near the battery compartment (because the battery energy consumption of logistics vehicles is relatively large, the temperature changes in the compartment are more significant), the humidity sensor simultaneously records the air humidity in the transportation area (such as the impact of high humidity environment in coastal areas on the circuit system), and the pressure sensor combines the altitude data obtained by the GPS positioning module (such as air pressure changes during transportation in mountainous areas). The GPS positioning module also simultaneously records the vehicle's driving areas (such as industrial parks, highways) and specific driving times on different transportation routes (distinguishing between high-frequency transportation periods during the day and parking periods at night).

[0050] The historical fault record data is stored in the local database of the electronic control unit of each vehicle. The data acquisition module reads the data in real time in batches through the cloud interface of the fleet management system. The dynamic weighted filtering algorithm is used to pre-process the historical fault data of common logistics vehicles such as battery life attenuation and motor overheating. The formula is: In this process, the weights are adjusted dynamically. For example, when the battery temperature data of a vehicle fluctuates abnormally, the system adjusts the weights according to the Calculate the current weight, if (Current fluctuation range) exceeds (preset threshold), then increase the current data weight , reduce the impact of historical data and quickly filter out sudden interference signals.

[0051] The data transmission module uses 5G technology to encapsulate each vehicle's pre-processed driving habit data (such as heavy-load acceleration characteristics), environmental data (such as battery compartment temperature curve) and historical fault data into a data packet containing the vehicle's unique identifier according to the TCP / IP protocol, and transmits it in real time to the automobile company's cloud server cluster through the fleet's dedicated virtual private network. The cloud-based data analysis and processing module performs layered analysis based on the characteristics of the logistics vehicle queue.

[0052] In the construction of the driving habit feature model, clustering algorithm is used to process the driving data of all vehicles in the fleet in batches, and the driving style quantitative evaluation formula is used to , calculate the quantitative score of each vehicle , since the logistics vehicle driving scene focuses more on economy, the weight coefficient The proportion of (acceleration characteristics) is relatively reduced, (braking characteristics) and (Steering characteristics) Dynamically adjust based on the characteristics of high-frequency braking and heavy-load steering in logistics transportation, ultimately dividing driving styles into sub-types such as "heavy-load economical type" and "urban frequent start-stop type", and establish a corresponding characteristic parameter library.

[0053] When constructing the vehicle usage environment characteristic model, the association rule algorithm is used to analyze the relationship between the logistics vehicle transportation route and environmental parameters. For example, it was found that in the high temperature and high humidity environment in a certain area in summer, the failure probability of the logistics vehicle battery management system was strongly correlated with the duration that the battery compartment temperature exceeded 40°C. Based on this, an environmental impact factor model was established. The historical fault association model used a time series analysis algorithm to explore the temporal relationship between logistics vehicle battery failures and mileage and charging frequency. For example, it was found that battery capacity attenuation failures often occurred at nodes with a mileage of more than 50,000 kilometers and more than 300 fast charging times. Combined with the frequency of sudden acceleration in driving habits, multi-dimensional fault association rules were established.

[0054] The personalized warning model construction module is based on a deep learning framework and adopts a distributed LSTM network architecture for logistics vehicle fleets. It uses each vehicle's driving habit feature model (such as heavy-load acceleration mode), environmental feature model (such as high-temperature transportation scenario) and historical fault association model (such as battery attenuation law) as input, and the fleet's historical fault data (such as motor controller overheating failure) as training labels. By adjusting the learning rate and the number of hidden layer neurons (considering the large amount of logistics vehicle data, the hidden layer dimension is increased), a personalized warning model is generated for each vehicle. The dynamic prediction formula of fault probability is used during the training process. , for example, when predicting battery thermal runaway failure, Function focuses on integrating battery temperature historical data , current ambient temperature and rapid acceleration frequency , through the feature weight parameter Dynamically adjust the impact ratio of each factor.

[0055] The fault diagnosis and early warning module collects data from various systems of logistics vehicles in real time, such as the single-cell voltage of the battery management system and the current value of the motor controller. When the predicted probability of the battery temperature of a vehicle exceeds the early warning threshold, the system generates an early warning message including "the battery compartment temperature is abnormal, the possible cause is the failure of the cooling fan, and it is recommended to stop immediately to check the cooling system". The message is sent synchronously to the driver terminal and the fleet administrator platform through the 5G network in the form of fleet management APP push and SMS. At the same time, the fault diagnosis results (such as the specific location of the battery temperature abnormality and historical similar fault handling solutions) are automatically uploaded to the big data platform of the vehicle company's after-sales service center, providing data support for the preventive maintenance of batch logistics vehicles.

[0056] In summary, the application of this embodiment in the new energy commercial logistics vehicle fleet has fully verified the effectiveness of the system in the batch vehicle management scenario. The personalized warning model is combined with the distributed deep learning architecture to establish a fault prediction model adapted to the logistics and transportation scenario for each vehicle; the fault diagnosis and warning module realizes the full process automation from real-time monitoring to multi-level warning. This embodiment not only provides fleet operators with accurate fault warning services, helping them to arrange maintenance plans in advance and reduce downtime losses, but also collects a large amount of actual operating data of commercial vehicle models for vehicle companies, helping to improve the design of battery thermal management systems and motor control systems. Compared with traditional general warning systems, this solution has significantly improved the fault warning accuracy in the logistics vehicle scenario, especially in the early fault identification of battery systems and motor systems, reducing the false alarm rate, and providing reliable technical support for the large-scale operation of new energy commercial vehicles.

[0057] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as above in terms of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can, without departing from the scope of the technical solution of the present invention, make some changes or modifications to equivalent embodiments using the technical contents disclosed above. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A new energy vehicle intelligent network remote diagnosis and fault warning system, characterized in that: The system includes: data acquisition module, data transmission module, data analysis and processing module, personalized early warning model construction module and fault diagnosis and early warning module; The data acquisition module is used to collect driving habit data, vehicle usage environment data and historical fault record data of new energy vehicles; The data transmission module is used to transmit the data acquired by the data acquisition module to the cloud server in real time; The data analysis and processing module is deployed on the cloud server and is used to analyze the transmitted data to obtain a driving habit feature model, a vehicle usage environment feature model, and a historical fault correlation model; The personalized warning model building module establishes a personalized fault warning model for each vehicle based on the driving habit feature model, the vehicle usage environment feature model, and the historical fault association model; The fault diagnosis and warning module monitors the operating status data of various vehicle systems and components in real time, inputs it into the corresponding personalized warning model for analysis, and when the warning threshold is reached or exceeded, sends fault warning information to the owner's mobile terminal through the data transmission module, and feeds back the fault diagnosis results to the vehicle manufacturer's after-sales service center.

2. A new energy vehicle intelligent network remote diagnosis and fault warning system according to claim 1, characterized in that: In the data acquisition module, driving habit data is collected through displacement sensors and pressure sensors installed on the accelerator pedal, brake pedal, and steering wheel, including the accelerator pedal travel change rate, brake pedal pedaling force and frequency, and steering angle change rate. Vehicle usage environment data is collected by environmental sensors and GPS positioning modules. The environmental sensors include temperature sensors, humidity sensors, and air pressure sensors. The collected data includes ambient temperature, humidity, and altitude. The GPS positioning module obtains the vehicle's driving area and driving time.

3. A new energy vehicle intelligent network remote diagnosis and fault warning system according to claim 2, characterized in that: The historical fault record data is recorded by the vehicle electronic control unit and stored in the local database. The data acquisition module reads it in real time and performs preliminary preprocessing of filtering and noise reduction after acquisition. The preprocessing adopts a dynamic weighted filtering algorithm to dynamically adjust the weight to achieve adaptive filtering of abnormal data. The calculation formula is: ,in, For the Second filter output value, For the The original data collected, is the dynamic weight coefficient at the current moment, satisfying is the current data fluctuation range, is the preset fluctuation threshold, To adjust the parameters, is the historical data weight coefficient, is the historical data window size.

4. The new energy vehicle intelligent network remote diagnosis and fault warning system according to claim 1 is characterized in that: The data transmission module adopts 4G / 5G technology to encapsulate the driving habit data, vehicle usage environment data and historical fault record data processed by the data acquisition module into data packets according to the TCP / IP communication protocol and transmit them to the cloud server.

5. The new energy vehicle intelligent network remote diagnosis and fault warning system according to claim 1 is characterized in that: The data analysis and processing module uses a clustering algorithm to perform cluster analysis on driving habit data, classifying driving styles into aggressive, stable, and economical types, and calculates corresponding driving habit characteristic parameters to establish a driving habit characteristic model. An association rule algorithm is used to analyze vehicle usage environment data to establish a vehicle usage environment characteristic model. A time series analysis algorithm is used to analyze historical fault record data, and a historical fault association model is established by combining driving habits and usage environment data. When establishing the driving habit characteristic model, a driving style quantitative evaluation formula is used to quantitatively score driving behavior to classify driving style types. The formula is: ,in, Quantify the driving style score, for The positive rate of change of the accelerator pedal at any moment, for The brake pedal is pressed hard at all times. for The change in steering angle at any moment, is the data collection period, 、 、 is the weight coefficient, and .

6. The new energy vehicle intelligent network remote diagnosis and fault warning system according to claim 1 is characterized in that: The personalized warning model construction module is based on a deep learning framework and adopts a long short-term memory network to build a model. It uses data from a driving habit feature model, a vehicle usage environment feature model, and a historical fault association model as input, and uses actual fault data from various vehicle systems and components as training labels for training. By adjusting the learning rate and the number of hidden layer neurons, the model prediction error is minimized to obtain a personalized fault warning model.

7. According to the new energy vehicle intelligent network remote diagnosis and fault warning system of claim 6, the training model adopts a dynamic prediction formula for fault probability to achieve dynamic prediction of fault probability through multi-feature fusion, the formula is: in, for Always based on historical fault data , environmental data , driving habit data The probability of failure, is the feature weight parameter, is a joint feature function that integrates historical fault correlation, environmental influencing factors and driving habit characteristics. 、 、 For the Feature vector of the class fault.

8. The new energy vehicle intelligent network remote diagnosis and fault warning system according to claim 1 is characterized in that: The fault diagnosis and warning module obtains the operating status data of various vehicle systems and components from the vehicle electronic control unit in real time, inputs the data into the personalized warning model, and when the model predicts that the probability of a fault exceeding a set threshold, generates fault warning information including the fault type, possible cause, and recommended measures, and sends it to the owner's mobile terminal through the data transmission module, and uploads the fault diagnosis results to the vehicle manufacturer's after-sales service center database.

9. The new energy vehicle intelligent network remote diagnosis and fault warning system according to claim 1 is characterized in that: After the data acquisition module collects data, it performs preliminary pre-processing of filtering and noise reduction on the data, and then transmits it to the cloud server through the data transmission module. The fault warning information sent by the fault diagnosis and warning module is sent to the car owner's mobile terminal in the form of text messages and push notifications.

Images