Vehicle-mounted terminal fault diagnosis system and method for Internet of Vehicles

Abstract

The invention relates to a vehicle-mounted terminal fault diagnosis system and method for the Internet of Vehicles. The method comprises the steps of monitoring and collecting vehicle state data of a vehicle in real time through a sensor network of a vehicle-mounted terminal, including an engine power value, a vehicle speed value and a tire pressure value, and transmitting the vehicle state data to a cloud server through vehicle-mounted communication equipment based on the Internet of Vehicles, and a data processing and analysis algorithm is introduced into the cloud server to carry out local time sequence mode feature analysis and full time domain time sequence association feature extraction on the vehicle state data so as to obtain posterior vehicle operation state features, and a vehicle operation state label and a vehicle-mounted terminal fault state are determined based on the posterior features. Therefore, whether the vehicle-mounted terminal has a fault or not can be judged by comparing the state data of the vehicle with the preset parameters of the vehicle-mounted terminal, and processing and maintenance can be carried out in time.

Description

Technical Field

[0001] The present disclosure relates to the technical field of Internet of Vehicles, and in particular, to a vehicle-mounted terminal fault diagnosis system and method for Internet of Vehicles. Background Art

[0002] The Internet of Vehicles refers to a network technology that connects vehicles to the Internet, roads, other vehicles, etc. through on-board communication equipment to achieve information interaction and sharing. The Internet of Vehicles can improve the safety, efficiency and comfort of vehicles, and also provide a new way to diagnose vehicle faults. The on-board terminal is the core component of the Internet of Vehicles. It is responsible for collecting, processing and transmitting vehicle status data, as well as receiving and executing remote control commands, providing key information for vehicle safety and performance. If the on-board terminal fails, it may affect the normal operation of the vehicle and even cause an accident. Therefore, it is very important to detect and handle the failure of the on-board terminal in a timely manner.

[0003] However, traditional vehicle terminal fault diagnosis systems mainly rely on experienced experts to judge faults by analyzing limited data and fault knowledge bases. This method is inefficient and inaccurate, and it is difficult to meet the requirements of fault diagnosis in the modern era of Internet of Vehicles. As the amount of data collected by vehicle terminals continues to increase, there is an urgent need for a vehicle terminal fault diagnosis system for Internet of Vehicles. Summary of the invention

[0004] This summary is provided to introduce concepts in a brief form that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0005] In a first aspect, the present disclosure provides a vehicle terminal fault diagnosis system for an Internet of Vehicles, the system comprising: A vehicle status data acquisition module, used to acquire a time series of vehicle status data of a detected vehicle collected by a sensor network of a vehicle-mounted terminal, wherein the vehicle status data includes an engine power value, a vehicle speed value, and a tire pressure value; A data transmission module, used to transmit the time series of the vehicle status data of the detected vehicle to a cloud server through an on-board communication device based on the Internet of Vehicles; A vehicle status data time series arrangement module, used for arranging the time series of the vehicle status data of the detected vehicle into a power time series input vector, a vehicle speed time series input vector and a tire pressure time series input vector according to the time dimension in the cloud server; A vehicle status data local time series feature analysis module, used for performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors; A vehicle status data full-time domain time series correlation feature extraction module, used for respectively passing the sequence of the power local time series feature vector, the sequence of the vehicle speed local time series feature vector and the sequence of the tire pressure local time series feature vector through a full-time domain time series correlation feature extractor based on an autocorrelation attention network to obtain a power time series correlation feature vector, a vehicle speed time series correlation feature vector and a tire pressure time series correlation feature vector; A vehicle running state characterization module, used for performing a fusion process based on a Bayesian probability model on the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector to obtain a posterior vehicle running state feature; The fault detection and early warning module is used to determine the operating status label of the detected vehicle based on the a posteriori vehicle operating status characteristics, and to determine whether there is a fault in the vehicle terminal and whether to generate a fault early warning prompt.

[0006] Optionally, the vehicle status data local time series feature analysis module includes: a vehicle status time series data segmentation unit, which is used to segment the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector based on a predetermined time scale to obtain a sequence of power local time series input vectors, a sequence of vehicle speed local time series input vectors and a sequence of tire pressure local time series input vectors; a state parameter local time series feature extraction unit, which is used to pass the sequence of power local time series input vectors, the sequence of vehicle speed local time series input vectors and the sequence of tire pressure local time series input vectors through a time series pattern feature extractor based on a one-dimensional extended convolutional network to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors.

[0007] Optionally, the state parameter local time series feature extraction unit is used to: process the sequence of the power local time series input vectors through the time series pattern feature extractor based on the one-dimensional extended convolutional network with the following feature extraction formula to obtain the sequence of the power local time series feature vectors; wherein the feature extraction formula is: ;in, respectively represent each power local timing input vector in the sequence of the power local timing input vector, A concatenated vector representing a sequence of power local timing input vectors, Indicates a connection operation. represents the sequence of the power local timing input vector The concatenated vector of and Represent the weight matrix and bias vector respectively, represents the convolution operation, The first The local time series feature vector of power.

[0008] Optionally, the vehicle status data full-time domain temporal correlation feature extraction module is used to: process the sequence of the power local temporal series feature vectors through the full-time domain temporal correlation feature extractor based on the self-correlation attention network with the following self-attention formula to obtain the power temporal correlation feature vector; wherein the self-attention formula is: ;in, is the first in the sequence of the power local time series feature vector The local time series characteristic vector of power, and Represent the weight coefficient vector and weight coefficient matrix respectively, is the offset vector, express function, For the The attention score value of the power local temporal feature vector, represents exponential operation, For the The attention score of the power local temporal feature vector, is the number of vectors in the sequence of the power local time series feature vector, For the The weight value of the power local time series feature vector, is the power timing associated feature vector.

[0009] Optionally, the vehicle operating state characterization module is used to: pass the power time series association feature vector, the vehicle speed time series association feature vector and the tire pressure time series association feature vector through a feature fusion network based on a Bayesian probability model to obtain a posterior vehicle operating state feature vector as the posterior vehicle operating state feature.

[0010] Optionally, the fault detection and early warning module includes: a vehicle state detection unit, used to pass the a posteriori vehicle operating state feature vector through a classifier-based vehicle operating state identifier to obtain an identification result, and the identification result is used to represent the operating state label of the detected vehicle; a fault detection unit, used to determine whether there is a fault in the vehicle-mounted terminal based on the operating state label of the detected vehicle and the actual operating state label of the detected vehicle; and a fault early warning unit, used to generate a fault early warning prompt in response to the vehicle-mounted terminal being faulty.

[0011] Optionally, it also includes a training module for training the temporal pattern feature extractor based on the one-dimensional extended convolutional network, the full-time domain temporal correlation feature extractor based on the autocorrelation attention network, the feature fusion network based on the Bayesian probability model, and the classifier-based vehicle fault detector.

[0012] Optionally, the training module includes: a training vehicle state data acquisition unit, which is used to obtain the time series of training vehicle state data of the detected vehicle collected by the sensor network of the vehicle-mounted terminal, wherein the training vehicle state data includes a training engine power value, a training vehicle speed value and a training tire pressure value; a training data transmission unit, which is used to transmit the time series of the training vehicle state data of the detected vehicle to a cloud server through an on-board communication device based on the Internet of Vehicles; and a training vehicle state data timing arrangement unit, which is used to transmit the time series of the training vehicle state data of the detected vehicle in the cloud server. The training power time series input vector, the training vehicle speed time series input vector and the training tire pressure time series input vector are arranged according to the time dimension; a local time series feature analysis unit for training vehicle status data is used to perform local time series pattern feature analysis on the training power time series input vector, the training vehicle speed time series input vector and the training tire pressure time series input vector to obtain a sequence of training power local time series feature vectors, a sequence of training vehicle speed local time series feature vectors and a sequence of training tire pressure local time series feature vectors; a full-time domain time series correlation feature extraction unit for training vehicle status data is used to extract the training power local time series feature vectors. The sequence of the feature vectors, the sequence of the training vehicle speed local time series feature vectors and the sequence of the training tire pressure local time series feature vectors are respectively passed through the full-time domain time series correlation feature extractor based on the autocorrelation attention network to obtain the training power time series correlation feature vector, the training vehicle speed time series correlation feature vector and the training tire pressure time series correlation feature vector; the training vehicle running state representation unit is used to pass the training power time series correlation feature vector, the training vehicle speed time series correlation feature vector and the training tire pressure time series correlation feature vector through the feature fusion network based on the Bayesian probability model to obtain the training post A posterior vehicle operating state feature vector is obtained; a training classification unit is used to pass the training a posteriori vehicle operating state feature vector through the classifier-based vehicle operating state identifier to obtain a classification loss function value; a training unit is used to train the temporal pattern feature extractor based on the one-dimensional extended convolutional network, the full-time domain temporal association feature extractor based on the autocorrelation attention network, the feature fusion network based on the Bayesian probability model and the vehicle fault detector based on the classifier based on the classification loss function value, and the training a posteriori vehicle operating state feature vector is optimized in each iterative training.

[0013] In a second aspect, the present disclosure provides a vehicle terminal fault diagnosis method for an Internet of Vehicles, the method comprising: Acquire a time series of vehicle status data of a detected vehicle collected by a sensor network of a vehicle-mounted terminal, wherein the vehicle status data includes an engine power value, a vehicle speed value, and a tire pressure value; Transmitting the time series of vehicle status data of the detected vehicle to a cloud server through an on-board communication device based on the Internet of Vehicles; In the cloud server, the time series of the vehicle status data of the detected vehicle is arranged according to the time dimension into a power time series input vector, a vehicle speed time series input vector and a tire pressure time series input vector; Performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors; The sequence of the power local time series feature vectors, the sequence of the vehicle speed local time series feature vectors and the sequence of the tire pressure local time series feature vectors are respectively passed through a full-time domain time series correlation feature extractor based on an autocorrelation attention network to obtain a power time series correlation feature vector, a vehicle speed time series correlation feature vector and a tire pressure time series correlation feature vector; Performing fusion processing based on a quasi-Bayesian probability model on the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector to obtain a posterior vehicle operating state feature; Based on the a posteriori vehicle operating status characteristics, the operating status label of the detected vehicle is determined, and it is determined whether there is a fault in the vehicle-mounted terminal and whether a fault warning prompt is generated.

[0014] Optionally, performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector, and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors, and a sequence of tire pressure local time series feature vectors includes: Based on a predetermined time scale, the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector are segmented to obtain a sequence of power partial time series input vectors, a sequence of vehicle speed partial time series input vectors and a sequence of tire pressure partial time series input vectors; The sequence of the power local time series input vectors, the sequence of the vehicle speed local time series input vectors and the sequence of the tire pressure local time series input vectors are respectively passed through a time series pattern feature extractor based on a one-dimensional extended convolutional network to obtain a sequence of the power local time series feature vectors, a sequence of the vehicle speed local time series feature vectors and a sequence of the tire pressure local time series feature vectors.

[0015] By adopting the above technical solution, the vehicle status data of the vehicle, including engine power value, vehicle speed value and tire pressure value, is collected in real time through the sensor network of the vehicle terminal, and these vehicle status data are transmitted to the cloud server through the vehicle communication equipment based on the Internet of Vehicles, so as to introduce data processing and analysis algorithms in the cloud server to perform local time series pattern feature analysis and full time domain time series correlation feature extraction on these vehicle status data, so as to obtain the a posteriori vehicle operation status feature, and determine the vehicle operation status label and the fault status of the vehicle terminal based on the a posteriori feature. In this way, by comparing the vehicle status data and the preset parameters of the vehicle terminal, it is possible to determine whether the vehicle terminal has a fault, and to handle and repair it in a timely manner.

[0016] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale. In the drawings: Figure 1 The present invention is a block diagram of a vehicle terminal fault diagnosis system for Internet of Vehicles according to an exemplary embodiment.

[0018] Figure 2 The present invention is a flowchart of a method for diagnosing faults of an on-board terminal for Internet of Vehicles according to an exemplary embodiment.

[0019] Figure 3 It is a block diagram of an electronic device according to an exemplary embodiment.

[0020] Figure 4 It is an application scenario diagram of an on-board terminal fault diagnosis system for Internet of Vehicles according to an exemplary embodiment. DETAILED DESCRIPTION

[0021] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0022] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0023] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0024] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0025] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0026] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0027] First, in response to the above technical problems, in the technical solution of this application, a vehicle-mounted terminal fault diagnosis system for the Internet of Vehicles is proposed, which can determine whether the vehicle-mounted terminal is working normally by analyzing the vehicle's status data. It should be understood that the vehicle's status data includes engine power value, vehicle speed value, tire pressure value, etc., which can reflect the vehicle's operating status and potential problems. For example, if the engine power value is lower than the normal range, it may indicate that there is a problem with the power supply of the vehicle-mounted terminal; if the vehicle speed value does not match the actual driving speed, it may indicate that the sensor or communication module of the vehicle-mounted terminal is faulty; if the tire pressure value is abnormal, it may indicate that there is a problem with the monitoring system or alarm system of the vehicle-mounted terminal. By comparing the vehicle's status data and the preset parameters of the vehicle-mounted terminal, it can be determined whether the vehicle-mounted terminal has a fault, and timely processing and maintenance can be carried out.

[0028] In order to solve the above problems, the present disclosure provides a vehicle terminal fault diagnosis system and method for the Internet of Vehicles, which collects vehicle status data of the vehicle in real time through the sensor network of the vehicle terminal, including engine power value, vehicle speed value and tire pressure value, and transmits these vehicle status data to the cloud server through the vehicle communication equipment based on the Internet of Vehicles, so as to introduce data processing and analysis algorithms in the cloud server to perform local time series pattern feature analysis and full time domain time series correlation feature extraction on these vehicle status data, so as to obtain the a posteriori vehicle operation status feature, and determine the vehicle operation status label and the fault status of the vehicle terminal based on the a posteriori feature. In this way, by comparing the vehicle status data and the preset parameters of the vehicle terminal, it can be determined whether the vehicle terminal has a fault, and timely processing and maintenance can be carried out.

[0029] The specific embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0030] Figure 1 FIG. 1 is a block diagram of a vehicle terminal fault diagnosis system for an Internet of Vehicles according to an exemplary embodiment. Figure 1 As shown, the vehicle terminal fault diagnosis system 100 for the Internet of Vehicles includes: The vehicle status data acquisition module 101 is used to acquire the time series of the vehicle status data of the detected vehicle collected by the sensor network of the vehicle terminal, wherein the vehicle status data includes the engine power value, the vehicle speed value and the tire pressure value; The data transmission module 102 is used to transmit the time series of the vehicle status data of the detected vehicle to the cloud server through the vehicle-mounted communication device based on the Internet of Vehicles; The vehicle status data time series arrangement module 103 is used to arrange the time series of the vehicle status data of the detected vehicle into a power time series input vector, a vehicle speed time series input vector and a tire pressure time series input vector according to the time dimension in the cloud server; The vehicle state data local time series feature analysis module 104 is used to perform local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors; The vehicle status data full-time domain temporal correlation feature extraction module 105 is used to respectively pass the sequence of the power local temporal series feature vector, the sequence of the vehicle speed local temporal series feature vector and the sequence of the tire pressure local temporal series feature vector through a full-time domain temporal correlation feature extractor based on an autocorrelation attention network to obtain a power temporal correlation feature vector, a vehicle speed temporal correlation feature vector and a tire pressure temporal correlation feature vector; The vehicle running state characterization module 106 is used to perform a fusion process based on a Bayesian probability model on the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector to obtain a posterior vehicle running state feature; The fault detection and warning module 107 is used to determine the running status label of the detected vehicle based on the a posteriori vehicle running status characteristics, and determine whether there is a fault in the vehicle terminal and whether to generate a fault warning prompt.

[0031] In particular, in order to be able to apply the correlation and timing change information between various status data of the vehicle, the technical concept of the present application is to monitor and collect the vehicle status data of the vehicle in real time through the sensor network of the vehicle-mounted terminal, including engine power value, vehicle speed value and tire pressure value, and transmit these vehicle status data to the cloud server through the vehicle-mounted communication equipment based on the Internet of Vehicles, so as to introduce data processing and analysis algorithms in the cloud server to perform timing coordinated analysis of these vehicle status data, thereby utilizing the timing coordinated correlation and changes between various vehicle operating status data to realize accurate identification of the vehicle operating status, and compare it with the actual operating status, so as to determine whether there is a fault in the vehicle-mounted terminal, so as to issue an early warning prompt in time when a fault is found.

[0032] Specifically, in the technical solution of the present application, first, the time series of vehicle status data of the detected vehicle collected by the sensor network of the vehicle terminal is obtained, wherein the vehicle status data includes engine power value, vehicle speed value and tire pressure value. It should be understood that the vehicle status data collected by the sensor network includes key parameters such as engine power value, vehicle speed value and tire pressure value, which reflect the operating status and performance of the vehicle. By performing time series analysis and feature extraction on these data, the operating status label of the vehicle can be detected more accurately, and compared with the actual operating status to discover the fault of the vehicle terminal, thereby improving the safety and reliability of the vehicle.

[0033] Next, considering that each vehicle status data of the detected vehicle is constantly changing with the passage of time, that is, each vehicle status data has a dynamic change rule in the time dimension, therefore, in order to analyze and feature mine the time series of the vehicle status data, it is necessary to transmit the time series of the vehicle status data of the detected vehicle to the cloud server through the vehicle-mounted communication equipment based on the Internet of Vehicles, and in the cloud server, arrange the time series of the vehicle status data of the detected vehicle according to the time dimension into a power time series input vector, a vehicle speed time series input vector and a tire pressure time series input vector.

[0034] In one embodiment of the present disclosure, the vehicle status data local time series feature analysis module includes: a vehicle status time series data segmentation unit, which is used to segment the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector based on a predetermined time scale to obtain a sequence of power local time series input vectors, a sequence of vehicle speed local time series input vectors and a sequence of tire pressure local time series input vectors; a state parameter local time series feature extraction unit, which is used to pass the sequence of power local time series input vectors, the sequence of vehicle speed local time series input vectors and the sequence of tire pressure local time series input vectors through a time series pattern feature extractor based on a one-dimensional extended convolutional network to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors.

[0035] It should be understood that, due to the large amount of data in the time series of the vehicle status data, directly processing long sequence data may bring huge computational complexity, and since the vehicle operation status often exhibits local patterns, that is, it has specific characteristics within a shorter time range. Therefore, in order to facilitate model learning and capture more detailed and comprehensive feature information of the vehicle status data in the time dimension, thereby improving the feature expression capability, in the technical solution of the present application, it is necessary to segment the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector based on a predetermined time scale to obtain a sequence of power local time series input vectors, a sequence of vehicle speed local time series input vectors and a sequence of tire pressure local time series input vectors.

[0036] Then, a one-dimensional convolution layer is used to perform feature analysis and extraction on the sequence of the power local time series input vector, the sequence of the vehicle speed local time series input vector, and the sequence of the tire pressure local time series input vector, respectively, so as to capture the local time series features of the engine power value, the vehicle speed value, and the tire pressure value in each local time period. In particular, considering that it is difficult to capture the time series correlation pattern and features under different step lengths due to the inherent limitations of one-dimensional convolution, therefore, in order to extract higher-level, more abstract and more comprehensive features from the time series of vehicle status data, so as to better understand and utilize the time series information of the vehicle status data, so as to better detect the vehicle running state, in the technical solution of the present application, the sequence of the power local time series input vector, the sequence of the vehicle speed local time series input vector, and the sequence of the tire pressure local time series input vector are respectively passed through a time series pattern feature extractor based on a one-dimensional extended convolutional network to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors, and a sequence of tire pressure local time series feature vectors. It should be understood that compared with the one-dimensional convolution layer, the one-dimensional expanded convolution layer introduces a hyperparameter called expansion rate to control the number of null values ​​0 in the convolution kernel. The expansion rate defines the spacing of the engine power value, vehicle speed value and tire pressure value in the time dimension at each time point when the convolution kernel processes the time series data. After the expanded convolution operation, the time series correlation characteristics of vehicle status data at different step lengths and at a long distance can be captured, providing a basis for subsequent time series correlation feature extraction and fault diagnosis.

[0037] In one embodiment of the present disclosure, the state parameter local time series feature extraction unit is used to: process the sequence of the power local time series input vector through the time series pattern feature extractor based on the one-dimensional extended convolutional network with the following feature extraction formula to obtain the sequence of the power local time series feature vector; wherein the feature extraction formula is: ;in, respectively represent each power local timing input vector in the sequence of the power local timing input vector, A concatenated vector representing a sequence of power local timing input vectors, Indicates a connection operation. represents the sequence of the power local timing input vector The concatenated vector of and Represent the weight matrix and bias vector respectively, represents the convolution operation, The first The local time series feature vector of power.

[0038] Furthermore, considering that for vehicle status data, such as engine power value, vehicle speed value and tire pressure value, the local time series patterns and features in each local time period in the time dimension have an association relationship based on the overall time series, and the local time series features of the vehicle status data in some local time periods play an important role in vehicle status recognition and detection, while some have a smaller role. Therefore, in order to be able to perform feature association encoding based on the full time domain for the local time series features of the vehicle status data in these local time periods, in the technical solution of the present application, the sequence of the power local time series feature vector, the sequence of the vehicle speed local time series feature vector and the sequence of the tire pressure local time series feature vector are further respectively passed through a full time domain time series association feature extractor based on an autocorrelation attention network to obtain a power time series association feature vector, a vehicle speed time series association feature vector and a tire pressure time series association feature vector. In particular, through the processing of the autocorrelation attention network, the correlation and dependency between different local time series features of various vehicle status data can be captured in the entire time domain, while important local time series features are weighted with high weights and less important local time series features are weighted with low weights. This is crucial for understanding the global pattern and correlation characteristics of the entire time series data, and helps to improve the accuracy and comprehensiveness of fault diagnosis.

[0039] In one embodiment of the present disclosure, the vehicle status data full-time domain temporal correlation feature extraction module is used to: process the sequence of the power local temporal series feature vectors through the full-time domain temporal correlation feature extractor based on the self-correlation attention network with the following self-attention formula to obtain the power temporal correlation feature vector; wherein the self-attention formula is: ;in, is the first in the sequence of the power local time series feature vector The local time series characteristic vector of power, and Represent the weight coefficient vector and weight coefficient matrix respectively, is the offset vector, express function, For the The attention score of the power local temporal feature vector, represents exponential operation, For the The attention score of the power local temporal feature vector, is the number of vectors in the sequence of the power local time series feature vector, For the The weight value of the power local time series feature vector, is the power timing associated feature vector.

[0040] Next, since the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector respectively represent the time series features of different aspects of the vehicle running state, in order to be able to more comprehensively and comprehensively represent the vehicle running state features, it is necessary to fuse the time series features of these vehicle running state data. Based on this, in the technical solution of the present application, the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector are further passed through a feature fusion network based on a Bayesian probability model to obtain a posterior vehicle running state feature vector. It should be understood that the Bayesian probability model can effectively handle the uncertainty of the time series features of various vehicle running state data, integrate the uncertainty information into the feature fusion process, and make the final posterior vehicle running state feature vector more reliable. In addition, through the feature fusion of the Bayesian probability model, the influence of the error of the time series correlation feature vector of the running state of a single vehicle on the final result can be reduced, and the noise and interference in the data can be better dealt with, and the robustness and stability of the detection system can be improved.

[0041] In one embodiment of the present disclosure, the vehicle operating state characterization module is used to: pass the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector through a feature fusion network based on a Bayesian probability model to obtain a posterior vehicle operating state feature vector as the posterior vehicle operating state feature.

[0042] Then, the a posteriori vehicle operating state feature vector is passed through a vehicle operating state identifier based on a classifier to obtain an identification result, and the identification result is used to represent the operating state label of the detected vehicle. Specifically, the classification label of the classifier is the operating state label of the detected vehicle. Therefore, after obtaining the classification result, the operating state of the vehicle can be more accurately judged and identified. Furthermore, based on the operating state label of the detected vehicle and the real operating state label of the detected vehicle, it is determined whether the vehicle-mounted terminal has a fault. In this way, the detected vehicle operating state and the real vehicle operating state can be compared to determine whether the vehicle-mounted terminal has a fault, and in response to the vehicle-mounted terminal having a fault, a fault warning prompt is generated. In this way, a more objective, accurate and intelligent vehicle terminal fault diagnosis can be achieved.

[0043] In one embodiment of the present disclosure, the fault detection and warning module includes: a vehicle state detection unit, used to pass the a posteriori vehicle operating state feature vector through a classifier-based vehicle operating state identifier to obtain an identification result, and the identification result is used to represent the operating state label of the detected vehicle; a fault detection unit, used to determine whether there is a fault in the vehicle terminal based on the operating state label of the detected vehicle and the actual operating state label of the detected vehicle; a fault warning unit, used to generate a fault warning prompt in response to the vehicle terminal being faulty.

[0044] In the technical solution of the present application, the sequence of local time series feature vectors of power, the sequence of local time series feature vectors of vehicle speed and the sequence of local time series feature vectors of tire pressure respectively express the local time series correlation pattern characteristics of the power value, the vehicle speed value and the tire pressure value in the local time domain. Therefore, after the sequence of local time series feature vectors of power, the sequence of local time series feature vectors of vehicle speed and the sequence of local time series feature vectors of tire pressure are respectively passed through the full time domain time series correlation feature extractor based on the autocorrelation attention network, the feature representation of each local time domain can be weighted and enhanced based on the overall distribution of time series features in the local time domain, that is, more focus can be placed on the time series feature distribution of certain local time domains.

[0045] However, considering the differences in the source time series distribution of the power value, the vehicle speed value and the tire pressure value, there will also be differences in the distribution of time series correlation features among the sequence of the power local time series feature vector, the sequence of the vehicle speed local time series feature vector and the sequence of the tire pressure local time series feature vector, and due to the reinforcement of certain local time series correlation features in the direction of time series distribution based on the autocorrelation attention network, the more significant time series correlation feature distributions among the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector are not aligned. In this way, when the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector are passed through a feature fusion network based on a Bayesian probability model, the posterior vehicle operation state feature vector obtained by the position-by-position Bayesian calculation of the feature vector will also have a significant problem of insufficient aggregation of the posterior time series correlation feature distribution, thereby affecting the convergence mapping iterative consistency of the posterior vehicle operation state feature vector to the representation of the predetermined class probability category in the probability space, and reducing the accuracy of the classification result.

[0046] In a preferred example, the a posteriori vehicle operating state feature vector is passed through a classifier-based vehicle operating state identifier to obtain a recognition result, comprising the following steps: optimizing the a posteriori vehicle operating state feature vector to obtain an optimized a posteriori vehicle operating state feature vector; and, passing the optimized a posteriori vehicle operating state feature vector through a classifier-based vehicle operating state identifier to obtain a recognition result.

[0047] The optimization of the a posteriori vehicle running state feature vector includes: optimizing the first feature value of the a posteriori vehicle running state feature vector and the second eigenvalue , calculate the absolute value of the difference divided by the sum of the absolute values ​​and then correct the relative probability value, wherein the correction makes the relative probability value greater than zero and less than one; calculate the first product of the first eigenvalue multiplied by the relative probability value, the second product of the second eigenvalue multiplied by the absolute value of the difference between the relative probability value and one, the first quotient of the first eigenvalue divided by the absolute value of the difference between the relative probability value and one, and the second quotient of the second eigenvalue divided by the relative probability value; after adding the first product and the second product, further subtract the difference between the first quotient and the second quotient to obtain the value equal to the first eigenvalue and the second eigenvalue The corresponding eigenvalues ​​of the corresponding optimization matrix ; and, multiplying the optimization matrix by the posterior vehicle operating state feature vector to obtain an optimized posterior vehicle operating state feature vector.

[0048] Specifically, it is expressed as: ;in, represents the posterior vehicle running state feature vector, and represent the first eigenvalue and the second eigenvalue of the posterior vehicle running state eigenvector respectively, represents the relative probability value, Represents the first eigenvalue and the second eigenvalue The corresponding eigenvalues ​​of the corresponding optimization matrix, Represents the first eigenvalue and the second eigenvalue The corresponding optimization matrix is, represents matrix multiplication, Represents the optimized posterior vehicle operating state feature vector.

[0049] Among them, the first eigenvalue and the second eigenvalue The corresponding relative probability value It is expressed as: ;in, and represent the first eigenvalue and the second eigenvalue of the posterior vehicle running state eigenvector respectively, represents the predetermined hyperparameters, Represents the relative probability value.

[0050] Here, with respect to the region-boundary integral relationship of the high-dimensional feature manifold of the posterior vehicle operating state feature vector, the probability distribution of the entire feature set of the posterior vehicle operating state feature vector is defined as a manifold constraint boundary to approximate the simply connected region representation of the high-dimensional manifold composed of eigenvalue pairs of the feature set of the posterior vehicle operating state feature vector, so as to avoid the ambiguity in the mapping of the diversified feature representation of the posterior vehicle operating state feature vector to the local manifold representation in the high-dimensional feature classification regression convergence space, improve the execution iteration consistency of each local feature distribution in the mapping task, improve the feature classification regression convergence effect, and improve the accuracy of the classification results.

[0051] In one embodiment of the present disclosure, the vehicle-mounted terminal fault diagnosis system for the Internet of Vehicles also includes a training module for training the temporal pattern feature extractor based on the one-dimensional extended convolutional network, the full-time domain temporal correlation feature extractor based on the autocorrelation attention network, the feature fusion network based on the Bayesian probability model, and the vehicle fault detector based on the classifier. The training module includes: a training vehicle status data acquisition unit, which is used to obtain the time series of training vehicle status data of the detected vehicle collected by the sensor network of the vehicle-mounted terminal, wherein the training vehicle status data includes training engine power values, training vehicle speed values, and training tire pressure values; a training data transmission unit, which is used to transmit the time series of the training vehicle status data of the detected vehicle to the cloud server through the vehicle-mounted communication device based on the Internet of Vehicles; a training vehicle status data time series arrangement unit, which is used to arrange the time series of the training vehicle status data of the detected vehicle in the cloud server according to the time series. The time-domain dimensions are arranged as a training power time series input vector, a training vehicle speed time series input vector, and a training tire pressure time series input vector; a training vehicle state data local time series feature analysis unit is used to perform local time series pattern feature analysis on the training power time series input vector, the training vehicle speed time series input vector, and the training tire pressure time series input vector to obtain a sequence of training power local time series feature vectors, a sequence of training vehicle speed local time series feature vectors, and a sequence of training tire pressure local time series feature vectors; a training vehicle state data full time domain time series correlation feature extraction unit is used to extract the training power local time series feature vectors into a local time series feature vector. The sequence of the training vehicle speed local time series feature vectors and the sequence of the training tire pressure local time series feature vectors are respectively passed through the full-time domain time series correlation feature extractor based on the autocorrelation attention network to obtain the training power time series correlation feature vector, the training vehicle speed time series correlation feature vector and the training tire pressure time series correlation feature vector; the training vehicle running state representation unit is used to pass the training power time series correlation feature vector, the training vehicle speed time series correlation feature vector and the training tire pressure time series correlation feature vector through the feature fusion network based on the Bayesian probability model to obtain the training posterior A vehicle operating state feature vector; a training classification unit, used to pass the training a posteriori vehicle operating state feature vector through the classifier-based vehicle operating state identifier to obtain a classification loss function value; a training unit, used to train the temporal pattern feature extractor based on the one-dimensional extended convolutional network, the full-time domain temporal association feature extractor based on the autocorrelation attention network, the feature fusion network based on the Bayesian probability model and the classifier-based vehicle fault detector based on the classification loss function value, and optimize the training a posteriori vehicle operating state feature vector in each iterative training.

[0052] In the technical solution of the present application, the sequence of training power local time series feature vectors, the sequence of training vehicle speed local time series feature vectors and the sequence of training tire pressure local time series feature vectors respectively express the local time series correlation pattern characteristics of the training engine power value, the training vehicle speed value and the training tire pressure value in the local time domain. Therefore, after the sequence of training power local time series feature vectors, the sequence of training vehicle speed local time series feature vectors and the sequence of training tire pressure local time series feature vectors are respectively passed through the full time domain time series correlation feature extractor based on the autocorrelation attention network, the feature representation of each local time domain can be weighted and enhanced based on the overall distribution of time series features in the local time domain, that is, more focus can be placed on the time series feature distribution of certain local time domains.

[0053] However, considering the differences in the source time series distribution of the training engine power value, the training vehicle speed value and the training tire pressure value, there will also be differences in the distribution of time series correlation features among the sequence of the training power local time series feature vector, the sequence of the training vehicle speed local time series feature vector and the sequence of the training tire pressure local time series feature vector, and due to the reinforcement of certain local time series correlation features in the direction of time series distribution based on the autocorrelation attention network, the more significant time series correlation feature distributions among the training power time series correlation feature vector, the training vehicle speed time series correlation feature vector and the training tire pressure time series correlation feature vector are not aligned. In this way, when the training power time series correlation feature vector, the training vehicle speed time series correlation feature vector and the training tire pressure time series correlation feature vector are passed through a feature fusion network based on a Bayesian probability model, the posterior vehicle running state feature vector obtained by the position-by-position Bayesian calculation of the feature vector will also have a significant problem of insufficient aggregation of the posterior time series correlation feature distribution, thereby affecting the classification convergence effect of the training posterior vehicle running state feature vector through the classifier, that is, affecting the classification training speed and the accuracy of the classification results. Therefore, the present application preferably performs the training of the posterior vehicle running state feature vector at each iterative training of the classifier. Make optimizations.

[0054] Specifically, during each iterative training, the training a posteriori vehicle running state feature vector is optimized, including: optimizing the training a posteriori vehicle running state feature vector The square of the eigenvalue of each position is related to the training posterior vehicle operation state feature vector The length of the training posterior vehicle operation state feature vector After adding the reciprocal of the second norm, the sum is used as the exponent of the exponential function with the natural constant as the base, and then the training posterior vehicle running state feature vector The square of the eigenvalue of each position is related to the training posterior vehicle operation state feature vector After multiplying the square of the norm of and the weighted hyperparameter, the product value is added to the above exponential value to obtain the optimized training posterior vehicle operation state feature vector .

[0055] In this way, by using the training posterior vehicle running state feature vector The low-rank norm representations of different orders are used as the posterior vehicle operation state feature vectors for the training The local space transformation decomposition coordinates of each eigenvalue of the training posterior vehicle running state eigenvector The overall distribution of the vector represents the spatial transformation of different ranks of the eigenvalues ​​to perform edge-based direction and distance prediction constraints on each eigenvalue as the center of the space, and the eigenvalues ​​belonging to the same regression class are mapped to a similar local spatial transformation decomposition coordinate set by aggregating the regression representation of the direction and scale of the feature distribution, thereby improving the training posterior vehicle operation state feature vector The aggregation effect of the feature set improves the training posterior vehicle running state feature vector Through the classification convergence effect of the classifier, the classification training speed and the accuracy of the classification results are improved. In this way, the time series synergistic correlation and change between various vehicle operation status data can be used to accurately identify the vehicle operation status and compare it with the actual operation status, so as to more accurately determine whether there is a fault in the vehicle terminal.

[0056] To sum up, the above scheme is adopted, which monitors and collects the vehicle status data of the vehicle in real time through the sensor network of the vehicle-mounted terminal, including the engine power value, vehicle speed value and tire pressure value, and transmits these vehicle status data to the cloud server through the vehicle-mounted communication equipment based on the Internet of Vehicles, so as to introduce data processing and analysis algorithms in the cloud server to perform time-series collaborative analysis of these vehicle status data, so as to utilize the time-series collaborative correlation and changes between multiple vehicle operation status data to realize accurate identification of the vehicle operation status, and compare it with the actual operation status, so as to determine whether there is a fault in the vehicle-mounted terminal, so as to issue a warning prompt in time when a fault is found.

[0057] Figure 2 is a flow chart of a vehicle terminal fault diagnosis method for vehicle networking according to an exemplary embodiment. Figure 2 As shown, the method includes: Step 201: Acquire a time series of vehicle status data of a detected vehicle collected by a sensor network of a vehicle-mounted terminal, wherein the vehicle status data includes an engine power value, a vehicle speed value, and a tire pressure value; Step 202: Transmitting the time series of the vehicle status data of the detected vehicle to a cloud server through a vehicle-mounted communication device based on the Internet of Vehicles; Step 203: In the cloud server, the time series of the vehicle status data of the detected vehicle is arranged according to the time dimension into a power time series input vector, a vehicle speed time series input vector and a tire pressure time series input vector; Step 204: performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors; Step 205: respectively pass the sequence of the power local time series feature vectors, the sequence of the vehicle speed local time series feature vectors, and the sequence of the tire pressure local time series feature vectors through a full-time domain time series correlation feature extractor based on an autocorrelation attention network to obtain a power time series correlation feature vector, a vehicle speed time series correlation feature vector, and a tire pressure time series correlation feature vector; Step 206: performing fusion processing based on a Bayesian probability model on the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector to obtain a posterior vehicle operating state feature; Step 207: Based on the a posteriori vehicle operating status characteristics, determine the operating status label of the detected vehicle, and determine whether there is a fault in the vehicle terminal and whether a fault warning prompt is generated.

[0058] Reference below Figure 3 , which shows a schematic diagram of the structure of an electronic device 600 suitable for implementing the embodiment of the present disclosure. The terminal device in the embodiment of the present disclosure may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0059] like Figure 3As shown, the electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0060] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 3 The electronic device 600 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0061] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0062] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0063] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0064] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0065] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0066] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0067] The modules involved in the embodiments described in the present disclosure may be implemented by software or hardware. The name of a module does not limit the module itself in some cases. For example, a test parameter acquisition module may also be described as a "module for acquiring device test parameters corresponding to a target device".

[0068] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0069] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0070] According to one or more embodiments of the present disclosure, Example 1 provides a vehicle terminal fault diagnosis system for an Internet of Vehicles, the system comprising: A vehicle status data acquisition module, used to acquire a time series of vehicle status data of a detected vehicle collected by a sensor network of a vehicle-mounted terminal, wherein the vehicle status data includes an engine power value, a vehicle speed value, and a tire pressure value; A data transmission module, used to transmit the time series of the vehicle status data of the detected vehicle to a cloud server through an on-board communication device based on the Internet of Vehicles; A vehicle status data time series arrangement module, used for arranging the time series of the vehicle status data of the detected vehicle into a power time series input vector, a vehicle speed time series input vector and a tire pressure time series input vector according to the time dimension in the cloud server; A vehicle status data local time series feature analysis module, used for performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors; A vehicle status data full-time domain time series correlation feature extraction module, used for respectively passing the sequence of the power local time series feature vector, the sequence of the vehicle speed local time series feature vector and the sequence of the tire pressure local time series feature vector through a full-time domain time series correlation feature extractor based on an autocorrelation attention network to obtain a power time series correlation feature vector, a vehicle speed time series correlation feature vector and a tire pressure time series correlation feature vector; A vehicle running state characterization module, used for performing a fusion process based on a Bayesian probability model on the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector to obtain a posterior vehicle running state feature; The fault detection and early warning module is used to determine the operating status label of the detected vehicle based on the a posteriori vehicle operating status characteristics, and to determine whether there is a fault in the vehicle terminal and whether to generate a fault early warning prompt.

[0071] According to one or more embodiments of the present disclosure, Example 2 provides the system of Example 1, wherein the vehicle status data local time series feature analysis module comprises: a vehicle state time series data segmentation unit, configured to segment the power time series input vector, the vehicle speed time series input vector, and the tire pressure time series input vector based on a predetermined time scale to obtain a sequence of power partial time series input vectors, a sequence of vehicle speed partial time series input vectors, and a sequence of tire pressure partial time series input vectors; The state parameter local time series feature extraction unit is used to pass the sequence of the power local time series input vector, the sequence of the vehicle speed local time series input vector and the sequence of the tire pressure local time series input vector through a time series pattern feature extractor based on a one-dimensional extended convolutional network to obtain the sequence of the power local time series feature vector, the sequence of the vehicle speed local time series feature vector and the sequence of the tire pressure local time series feature vector.

[0072] According to one or more embodiments of the present disclosure, Example 3 provides the system of Example 2, wherein the state parameter local time series feature extraction unit is used to: process the sequence of the power local time series input vectors through the time series pattern feature extractor based on the one-dimensional extended convolutional network with the following feature extraction formula to obtain the sequence of the power local time series feature vectors; Among them, the feature extraction formula is: ;in, respectively represent each power local timing input vector in the sequence of the power local timing input vector, A concatenated vector representing a sequence of power local timing input vectors, Indicates a connection operation. represents the sequence of the power local timing input vector The concatenated vector of and denote the weight matrix and the offset vector respectively, represents the convolution operation, The first The local time series feature vector of power.

[0073] According to one or more embodiments of the present disclosure, Example 4 provides the system of Example 3, wherein the vehicle status data full-time domain temporal correlation feature extraction module is used to: process the sequence of the power local temporal series feature vectors through the full-time domain temporal correlation feature extractor based on the autocorrelation attention network with the following self-attention formula to obtain the power temporal correlation feature vector; Among them, the self-attention formula is: ;in, is the first in the sequence of the power local time series feature vector The local time series characteristic vector of power, and Represent the weight coefficient vector and weight coefficient matrix respectively, is the offset vector, express function, For the The attention score of the power local temporal feature vector, represents exponential operation, For the The attention score of the power local temporal feature vector, is the number of vectors in the sequence of the power local time series feature vector, For the The weight value of the power local time series feature vector, is the power timing associated feature vector.

[0074] According to one or more embodiments of the present disclosure, Example 5 provides the system of Example 4, wherein the vehicle operating state characterization module is used to: pass the power timing association feature vector, the vehicle speed timing association feature vector and the tire pressure timing association feature vector through a feature fusion network based on a Bayesian probability model to obtain a posterior vehicle operating state feature vector as the posterior vehicle operating state feature.

[0075] According to one or more embodiments of the present disclosure, Example 6 provides the system of Example 5, wherein the fault detection and early warning module comprises: A vehicle state detection unit, used for passing the a posteriori vehicle running state feature vector through a vehicle running state identifier based on a classifier to obtain a recognition result, wherein the recognition result is used to represent a running state label of the detected vehicle; A fault detection unit, used to determine whether there is a fault in the vehicle terminal based on the running status label of the detected vehicle and the real running status label of the detected vehicle; The fault warning unit is used to generate a fault warning prompt in response to the vehicle-mounted terminal having a fault.

[0076] According to one or more embodiments of the present disclosure, Example 7 provides the system of Example 6, further including a training module for training the temporal pattern feature extractor based on the one-dimensional extended convolutional network, the full-time domain temporal correlation feature extractor based on the autocorrelation attention network, the feature fusion network based on the Bayesian probability model, and the classifier-based vehicle fault detector.

[0077] According to one or more embodiments of the present disclosure, Example 8 provides the system of Example 7, wherein the training module includes: A training vehicle state data acquisition unit, used to acquire a time series of training vehicle state data of a detected vehicle acquired by a sensor network of a vehicle-mounted terminal, wherein the training vehicle state data includes a training engine power value, a training vehicle speed value, and a training tire pressure value; A training data transmission unit, used to transmit the time series of the training vehicle state data of the detected vehicle to a cloud server via an on-board communication device based on the Internet of Vehicles; A training vehicle state data time series arrangement unit, used for arranging the time series of the training vehicle state data of the detected vehicle into a training power time series input vector, a training vehicle speed time series input vector and a training tire pressure time series input vector according to the time dimension in the cloud server; A training vehicle state data local time series feature analysis unit, used for performing local time series pattern feature analysis on the training power time series input vector, the training vehicle speed time series input vector and the training tire pressure time series input vector to obtain a sequence of training power local time series feature vectors, a sequence of training vehicle speed local time series feature vectors and a sequence of training tire pressure local time series feature vectors; A training vehicle state data full-time domain time series correlation feature extraction unit, used for respectively passing the sequence of the training power local time series feature vectors, the sequence of the training vehicle speed local time series feature vectors and the sequence of the training tire pressure local time series feature vectors through the full-time domain time series correlation feature extractor based on the autocorrelation attention network to obtain a training power time series correlation feature vector, a training vehicle speed time series correlation feature vector and a training tire pressure time series correlation feature vector; A training vehicle running state characterization unit, used for obtaining a training posterior vehicle running state feature vector by passing the training power time series associated feature vector, the training vehicle speed time series associated feature vector and the training tire pressure time series associated feature vector through the feature fusion network based on the quasi-Bayesian probability model; A training classification unit, used for passing the training posterior vehicle running state feature vector through the classifier-based vehicle running state identifier to obtain a classification loss function value; A training unit is used to train the temporal pattern feature extractor based on the one-dimensional extended convolutional network, the full-time domain temporal association feature extractor based on the autocorrelation attention network, the feature fusion network based on the Bayesian probability model and the vehicle fault detector based on the classifier based on the classification loss function value, and optimize the training posterior vehicle operation state feature vector in each iterative training.

[0078] According to one or more embodiments of the present disclosure, Example 9 provides a vehicle terminal fault diagnosis method for an Internet of Vehicles, the method comprising: Acquire a time series of vehicle status data of a detected vehicle collected by a sensor network of a vehicle-mounted terminal, wherein the vehicle status data includes an engine power value, a vehicle speed value, and a tire pressure value; Transmitting the time series of vehicle status data of the detected vehicle to a cloud server through an on-board communication device based on the Internet of Vehicles; In the cloud server, the time series of the vehicle status data of the detected vehicle is arranged according to the time dimension into a power time series input vector, a vehicle speed time series input vector and a tire pressure time series input vector; Performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors; The sequence of the power local time series feature vectors, the sequence of the vehicle speed local time series feature vectors and the sequence of the tire pressure local time series feature vectors are respectively passed through a full-time domain time series correlation feature extractor based on an autocorrelation attention network to obtain a power time series correlation feature vector, a vehicle speed time series correlation feature vector and a tire pressure time series correlation feature vector; Performing fusion processing based on a quasi-Bayesian probability model on the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector to obtain a posterior vehicle operating state feature; Based on the a posteriori vehicle operating status characteristics, the operating status label of the detected vehicle is determined, and it is determined whether there is a fault in the vehicle-mounted terminal and whether a fault warning prompt is generated.

[0079] According to one or more embodiments of the present disclosure, Example 10 provides the method of Example 9, performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector, and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors, and a sequence of tire pressure local time series feature vectors, including: Based on a predetermined time scale, the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector are segmented to obtain a sequence of power partial time series input vectors, a sequence of vehicle speed partial time series input vectors and a sequence of tire pressure partial time series input vectors; The sequence of the power local time series input vectors, the sequence of the vehicle speed local time series input vectors and the sequence of the tire pressure local time series input vectors are respectively passed through a time series pattern feature extractor based on a one-dimensional extended convolutional network to obtain a sequence of the power local time series feature vectors, a sequence of the vehicle speed local time series feature vectors and a sequence of the tire pressure local time series feature vectors.

[0080] Figure 4 FIG. 1 is an application scenario diagram of a vehicle terminal fault diagnosis system for an Internet of Vehicles according to an exemplary embodiment. Figure 4 As shown, in this application scenario, first, a time series of vehicle status data of a detected vehicle collected by a sensor network of a vehicle-mounted terminal is obtained, wherein the vehicle status data includes an engine power value (for example, Figure 4 C1 as shown in the figure), vehicle speed value (for example, Figure 4 C2) and tire pressure values ​​(e.g. Figure 4 Then, the obtained engine power value, vehicle speed value and tire pressure value are input to a server (for example, Figure 4 In S) shown in , the server can process the engine power value, the vehicle speed value and the tire pressure value based on the vehicle terminal fault diagnosis algorithm for the Internet of Vehicles to determine the label value of the gongche notation symbol.

[0081] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

[0082] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0083] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims. Regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.

Claims

1. A vehicle terminal fault diagnosis system for Internet of Vehicles, characterized in that: include: A vehicle status data acquisition module, used to acquire a time series of vehicle status data of a detected vehicle collected by a sensor network of a vehicle-mounted terminal, wherein the vehicle status data includes an engine power value, a vehicle speed value, and a tire pressure value; A data transmission module, used to transmit the time series of the vehicle status data of the detected vehicle to a cloud server through an on-board communication device based on the Internet of Vehicles; A vehicle status data time series arrangement module, used for arranging the time series of the vehicle status data of the detected vehicle into a power time series input vector, a vehicle speed time series input vector and a tire pressure time series input vector according to the time dimension in the cloud server; A vehicle status data local time series feature analysis module, used for performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors; A vehicle status data full-time domain time series correlation feature extraction module, used for respectively passing the sequence of the power local time series feature vector, the sequence of the vehicle speed local time series feature vector and the sequence of the tire pressure local time series feature vector through a full-time domain time series correlation feature extractor based on an autocorrelation attention network to obtain a power time series correlation feature vector, a vehicle speed time series correlation feature vector and a tire pressure time series correlation feature vector; A vehicle running state characterization module, used for performing a fusion process based on a Bayesian probability model on the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector to obtain a posterior vehicle running state feature; The fault detection and early warning module is used to determine the operating status label of the detected vehicle based on the a posteriori vehicle operating status characteristics, and to determine whether there is a fault in the vehicle terminal and whether to generate a fault early warning prompt.

2. The vehicle terminal fault diagnosis system for the Internet of Vehicles according to claim 1 is characterized in that: The vehicle status data local time series feature analysis module includes: a vehicle state time series data segmentation unit, configured to segment the power time series input vector, the vehicle speed time series input vector, and the tire pressure time series input vector based on a predetermined time scale to obtain a sequence of power partial time series input vectors, a sequence of vehicle speed partial time series input vectors, and a sequence of tire pressure partial time series input vectors; The state parameter local time series feature extraction unit is used to pass the sequence of the power local time series input vector, the sequence of the vehicle speed local time series input vector and the sequence of the tire pressure local time series input vector through a time series pattern feature extractor based on a one-dimensional extended convolutional network to obtain the sequence of the power local time series feature vector, the sequence of the vehicle speed local time series feature vector and the sequence of the tire pressure local time series feature vector.

3. The vehicle terminal fault diagnosis system for Internet of Vehicles according to claim 2 is characterized in that: The state parameter local time series feature extraction unit is used to: process the sequence of the power local time series input vectors through the time series pattern feature extractor based on the one-dimensional extended convolutional network with the following feature extraction formula to obtain the sequence of the power local time series feature vectors; Among them, the feature extraction formula is: ;in, respectively represent each power local timing input vector in the sequence of the power local timing input vector, A concatenated vector representing a sequence of power local timing input vectors, Indicates a connection operation. represents the sequence of the power local timing input vector The concatenated vector of and denote the weight matrix and the offset vector respectively, represents the convolution operation, The first The local time series feature vector of power.

4. The vehicle terminal fault diagnosis system for Internet of Vehicles according to claim 3 is characterized in that: The vehicle status data full-time domain time series correlation feature extraction module is used to: process the sequence of the power local time series feature vectors through the full-time domain time series correlation feature extractor based on the autocorrelation attention network with the following self-attention formula to obtain the power time series correlation feature vector; Among them, the self-attention formula is: ;in, is the first in the sequence of the power local time series feature vector The local time series characteristic vector of power, and Represent the weight coefficient vector and weight coefficient matrix respectively, is the offset vector, express function, For the The attention score of the power local temporal feature vector, represents exponential operation, For the The attention score of the power local temporal feature vector, is the number of vectors in the sequence of the power local time series feature vector, For the The weight value of the power local time series feature vector, is the power timing associated feature vector.

5. The vehicle-mounted terminal fault diagnosis system for Internet of Vehicles according to claim 4 is characterized in that: The vehicle operating state characterization module is used to: pass the power time series associated feature vector, the vehicle speed time series associated feature vector and the tire pressure time series associated feature vector through a feature fusion network based on a Bayesian probability model to obtain a posterior vehicle operating state feature vector as the posterior vehicle operating state feature.

6. The vehicle terminal fault diagnosis system for Internet of Vehicles according to claim 5, characterized in that: The fault detection and early warning module comprises: A vehicle state detection unit, used for passing the a posteriori vehicle running state feature vector through a vehicle running state identifier based on a classifier to obtain a recognition result, wherein the recognition result is used to represent a running state label of the detected vehicle; A fault detection unit, used to determine whether there is a fault in the vehicle terminal based on the running status label of the detected vehicle and the real running status label of the detected vehicle; The fault warning unit is used to generate a fault warning prompt in response to the vehicle-mounted terminal having a fault.

7. The vehicle terminal fault diagnosis system for Internet of Vehicles according to claim 6 is characterized in that: It also includes a training module for training the temporal pattern feature extractor based on the one-dimensional extended convolutional network, the full-time domain temporal correlation feature extractor based on the autocorrelation attention network, the feature fusion network based on the Bayesian probability model, and the vehicle fault detector based on the classifier.

8. The vehicle terminal fault diagnosis system for Internet of Vehicles according to claim 7, characterized in that: The training module comprises: A training vehicle state data acquisition unit, used to acquire a time series of training vehicle state data of a detected vehicle acquired by a sensor network of a vehicle-mounted terminal, wherein the training vehicle state data includes a training engine power value, a training vehicle speed value, and a training tire pressure value; A training data transmission unit, used to transmit the time series of the training vehicle state data of the detected vehicle to a cloud server via an on-board communication device based on the Internet of Vehicles; A training vehicle state data time series arrangement unit, used for arranging the time series of the training vehicle state data of the detected vehicle into a training power time series input vector, a training vehicle speed time series input vector and a training tire pressure time series input vector according to the time dimension in the cloud server; A training vehicle state data local time series feature analysis unit, used for performing local time series pattern feature analysis on the training power time series input vector, the training vehicle speed time series input vector and the training tire pressure time series input vector to obtain a sequence of training power local time series feature vectors, a sequence of training vehicle speed local time series feature vectors and a sequence of training tire pressure local time series feature vectors; A training vehicle state data full-time domain time series correlation feature extraction unit, used for respectively passing the sequence of the training power local time series feature vectors, the sequence of the training vehicle speed local time series feature vectors and the sequence of the training tire pressure local time series feature vectors through the full-time domain time series correlation feature extractor based on the autocorrelation attention network to obtain a training power time series correlation feature vector, a training vehicle speed time series correlation feature vector and a training tire pressure time series correlation feature vector; A training vehicle running state characterization unit, used for obtaining a training posterior vehicle running state feature vector by passing the training power time series associated feature vector, the training vehicle speed time series associated feature vector and the training tire pressure time series associated feature vector through the feature fusion network based on the quasi-Bayesian probability model; A training classification unit, used for passing the training posterior vehicle running state feature vector through the classifier-based vehicle running state identifier to obtain a classification loss function value; A training unit is used to train the temporal pattern feature extractor based on the one-dimensional extended convolutional network, the full-time domain temporal association feature extractor based on the autocorrelation attention network, the feature fusion network based on the Bayesian probability model and the vehicle fault detector based on the classifier based on the classification loss function value, and optimize the training posterior vehicle operation state feature vector in each iterative training.

9. A vehicle terminal fault diagnosis method for Internet of Vehicles, characterized in that: include: Acquire a time series of vehicle status data of a detected vehicle collected by a sensor network of a vehicle-mounted terminal, wherein the vehicle status data includes an engine power value, a vehicle speed value, and a tire pressure value; Transmitting the time series of vehicle status data of the detected vehicle to a cloud server through an on-board communication device based on the Internet of Vehicles; In the cloud server, the time series of the vehicle status data of the detected vehicle is arranged according to the time dimension into a power time series input vector, a vehicle speed time series input vector and a tire pressure time series input vector; Performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors and a sequence of tire pressure local time series feature vectors; The sequence of the power local time series feature vectors, the sequence of the vehicle speed local time series feature vectors and the sequence of the tire pressure local time series feature vectors are respectively passed through a full-time domain time series correlation feature extractor based on an autocorrelation attention network to obtain a power time series correlation feature vector, a vehicle speed time series correlation feature vector and a tire pressure time series correlation feature vector; Performing fusion processing based on a quasi-Bayesian probability model on the power time series correlation feature vector, the vehicle speed time series correlation feature vector and the tire pressure time series correlation feature vector to obtain a posterior vehicle operating state feature; Based on the a posteriori vehicle operating status characteristics, the operating status label of the detected vehicle is determined, and it is determined whether there is a fault in the vehicle-mounted terminal and whether a fault warning prompt is generated.

10. The vehicle terminal fault diagnosis method for Internet of Vehicles according to claim 9, characterized in that: Performing local time series pattern feature analysis on the power time series input vector, the vehicle speed time series input vector, and the tire pressure time series input vector to obtain a sequence of power local time series feature vectors, a sequence of vehicle speed local time series feature vectors, and a sequence of tire pressure local time series feature vectors, including: Based on a predetermined time scale, the power time series input vector, the vehicle speed time series input vector and the tire pressure time series input vector are segmented to obtain a sequence of power partial time series input vectors, a sequence of vehicle speed partial time series input vectors and a sequence of tire pressure partial time series input vectors; The sequence of the power local time series input vectors, the sequence of the vehicle speed local time series input vectors and the sequence of the tire pressure local time series input vectors are respectively passed through a time series pattern feature extractor based on a one-dimensional extended convolutional network to obtain a sequence of the power local time series feature vectors, a sequence of the vehicle speed local time series feature vectors and a sequence of the tire pressure local time series feature vectors.

Images