Vehicle signal field mapping method, electronic equipment and storage medium

Abstract

The invention provides a field mapping method for a vehicle signal, electronic equipment and a storage medium, and the method comprises the steps: obtaining a current vehicle signal of a to-be-mapped vehicle model, determining a signal type corresponding to the current vehicle signal, determining a target model in a plurality of field mapping models which are trained according to the signal type, and carrying out the mapping of the target model, the field information corresponding to the current vehicle signal is determined through the target model, signal field mapping based on the model is achieved, manual comparison of the vehicle signal matrix is not needed, the mapping efficiency and accuracy of the vehicle signal are greatly improved, and therefore the efficiency and accuracy of large-scale vehicle signal integration and analysis can be improved; according to the method, the field information is identified through the field mapping models corresponding to the signal types, the processing efficiency of the vehicle signals can be improved, the processing real-time performance of the vehicle signals is guaranteed, delay is reduced, the different signal characteristics of different signal types are considered, the field information is identified through the field mapping models corresponding to the signal types, and the field mapping precision can be further improved.

Description

Technical Field

[0001] The present application relates to the field of automobile technology, and in particular to a field mapping method, electronic device and storage medium for vehicle signals. Background Art

[0002] In the current technical field, traditional methods are generally used for the integration and analysis of vehicle signals, that is, relying on manual comparison of complex and diverse vehicle signal matrices, converting signal names and signal values ​​one by one to build a mapping relationship.

[0003] However, this method is not only time-consuming, labor-intensive and inefficient, but also highly dependent on manual operation. It is very easy to cause mapping errors due to human factors, which in turn leads to problems in the integration of vehicle signals. Therefore, existing technologies seriously restrict the ability to efficiently and accurately integrate and analyze large-scale vehicle signals. Summary of the invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, the present application aims to provide a field mapping method, electronic device and storage medium for vehicle signals, so as to solve the problems of low efficiency and low accuracy of vehicle signal mapping in the prior art, and to improve the efficiency and accuracy of vehicle signal integration and analysis.

[0005] An embodiment of the present application provides a field mapping method for a vehicle signal, the method comprising: Obtaining a current vehicle signal of a vehicle model to be mapped, and determining a signal type corresponding to the current vehicle signal; A target model is determined from among multiple trained field mapping models based on the signal type, and field information corresponding to the current vehicle signal is determined based on the target model.

[0006] Optionally, the multiple field mapping models include a convolutional neural network model and a pre-trained language representation model, and determining a target model from the multiple trained field mapping models based on the signal type includes: Determining whether the signal type is a text type; If so, the trained pre-trained language representation model is determined as the target model; otherwise, the trained convolutional neural network model is determined as the target model.

[0007] Optionally, the training process of the field mapping model includes: Collect sample vehicle signals of each vehicle type and determine key metadata of each sample vehicle signal; For each sample vehicle signal of a numerical type, determining a waveform feature of the sample vehicle signal based on key metadata of the sample vehicle signal, and for each sample vehicle signal of a text type, determining a semantic feature of the sample vehicle signal based on the key metadata of the sample vehicle signal; Based on the waveform features of each sample vehicle signal of numerical type and the corresponding field labels, a convolutional neural network is trained, and based on the semantic features of each sample vehicle signal of text type and the corresponding field labels, a pre-trained language representation network is trained to obtain the convolutional neural network model and the pre-trained language representation model.

[0008] Optionally, after obtaining the convolutional neural network model and the pre-trained language representation model, the method further includes: Obtain optimized training data set; Based on the vehicle signals of numerical type and the corresponding field labels in the optimized training data set, the convolutional neural network model is optimized and trained to update the convolutional neural network model; Based on the text-type vehicle signals and corresponding field labels in the optimized training data set, the pre-trained language representation model is optimized and trained to update the pre-trained language representation model.

[0009] Optionally, the method further includes: In response to detecting the vehicle model change information, acquiring a vehicle signal of a changed vehicle model corresponding to the vehicle model change information; At least one of the field mapping models is updated based on the vehicle signal of the changed vehicle model and the corresponding field information.

[0010] Optionally, after determining the field information corresponding to the current vehicle signal based on the target model, the method further includes: Displaying the current vehicle signal and corresponding field information in the signal display interface; In response to detecting that a user triggers a confirmation abnormality operation for the current vehicle signal in the signal display interface, the current vehicle signal is determined as a negative sample signal, and the corresponding field information is determined as a label corresponding to the negative sample signal, so as to update the target model based on the negative sample information and the corresponding label.

[0011] Optionally, after determining the field information corresponding to the current vehicle signal based on the target model, the method further includes: Detecting whether the field information exists in a preset field table, and whether the signal value content of the current vehicle signal satisfies a content rule corresponding to the field information in the preset field table; In response to the field information not existing in the preset field table, or the signal value content not satisfying the content rule, abnormal prompt information is displayed on the signal display interface, and an abnormal prompt signal is sent to the associated terminal device.

[0012] Optionally, after determining the field information corresponding to the current vehicle signal based on the target model, the method further includes: Performing format conversion on the current vehicle signal based on the field information to obtain a target signal, and associating and storing the target signal under the corresponding field information in a standardized database; Check whether there is a field in the standardized database that has not been associated with the target signal. If so, display abnormal prompt information on the signal display interface and send an abnormal prompt signal to the associated terminal device.

[0013] An embodiment of the present application further provides an electronic device, the electronic device comprising: Processor and memory; The processor is used to execute the steps of the field mapping method of the vehicle signal provided in any embodiment of the present application by calling the program or instruction stored in the memory.

[0014] An embodiment of the present application also provides a computer-readable storage medium, which stores a program or instruction, and the program or instruction enables a computer to execute the steps of the field mapping method of the vehicle signal provided in any embodiment of the present application.

[0015] In summary, the present application proposes a field mapping method for vehicle signals. The method obtains the current vehicle signal of the vehicle model to be mapped and determines the signal type corresponding to the current vehicle signal, and then determines the target model in the multiple field mapping models that have been trained according to the signal type. The field information corresponding to the current vehicle signal is determined by the target model, thereby realizing model-based signal field mapping. There is no need to manually compare the vehicle signal matrix, which greatly improves the mapping efficiency and accuracy of the vehicle signal, thereby improving the efficiency and accuracy of large-scale vehicle signal integration and analysis, and improving the processing efficiency of vehicle signals, ensuring the real-time processing of vehicle signals, and reducing delays. In addition, the method takes into account the different signal characteristics of different signal types, and identifies the field information by the field mapping model corresponding to the signal type, which can further improve the field mapping accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0017] Figure 1 is a flow chart of a field mapping method of a vehicle signal provided in an embodiment of the present application; Figure 2 is a schematic diagram of a field mapping system for a vehicle signal provided in an embodiment of the present application; Figure 3 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0019] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0020] As mentioned in the background technology, in response to the problems in the prior art, the present application proposes a field mapping method for vehicle signals. Figure 1 is a flow chart of a field mapping method of a vehicle signal provided by an embodiment of the present application. Figure 1 , the field mapping method of the vehicle signal specifically includes: S110: Acquire the current vehicle signal of the vehicle type to be mapped, and determine the signal type corresponding to the current vehicle signal.

[0021] The vehicle model to be mapped may be a vehicle model for which field mapping of vehicle signals is required, that is, a vehicle model for which field information corresponding to vehicle signals needs to be identified. The current vehicle signal may be a signal generated in real time by a vehicle of the vehicle model to be mapped.

[0022] Exemplarily, the current vehicle signal may be a signal describing relevant data during the vehicle's driving process, such as a vehicle speed signal, an acceleration signal, an accelerator pedal depth signal, etc.; or, the current vehicle signal may be a signal describing the status of each system during the vehicle's driving process, such as an intelligent driving system status signal, a vehicle controller status signal, etc.; or, the current vehicle signal may be a signal describing component failure during the vehicle's driving process, such as an engine failure signal, a headlight failure signal, etc.

[0023] Specifically, multi-protocol interfaces may be pre-integrated, such as CAN (Controller Area Network), LIN (Local Interconnect Network), Ethernet, etc., and current vehicle signals may be captured from the vehicle control systems of each vehicle model to be mapped through each interface.

[0024] In the embodiment of the present application, streaming data processing technology can also be introduced to obtain the current vehicle signal of the vehicle model to be mapped. Among them, streaming data processing can perform real-time processing on the continuously generated data stream. By obtaining the current vehicle signal of the vehicle model to be mapped through streaming data processing, it can be ensured that the vehicle signal is captured and processed at the same time as it is produced, reducing delays and improving response speed.

[0025] In an embodiment of the present application, taking into account the differences in signal characteristics between signals of different signal types, in order to improve the accuracy of field recognition of vehicle signals, it is also possible to first determine the corresponding signal type after acquiring the current vehicle signal, so as to select the corresponding field mapping model for the signal type to identify the field information.

[0026] The signal type may be the type of the signal value content in the current vehicle signal. Exemplarily, considering that numerical values ​​usually have waveform characteristics and texts usually have semantic characteristics, in the embodiment of the present application, the signal type may also be divided into numerical type and text type. Alternatively, considering that there is a difference in information contained in a signal with a large amount of data and a signal with a small amount of data, the signal type may be further divided into a large data type and a small data type.

[0027] Specifically, the signal value content in the current vehicle signal can be extracted, and the signal type corresponding to the current vehicle signal can be determined based on the signal value content; alternatively, a type flag can be pre-specified in the vehicle signal, and the type flag is used to describe the signal type. After acquiring the current vehicle signal, the corresponding signal type can be determined by the type flag in the current vehicle signal.

[0028] S120. Determine a target model from among the trained multiple field mapping models based on the signal type, and determine field information corresponding to the current vehicle signal based on the target model.

[0029] The field mapping model may be a pre-trained deep learning model. It should be noted that the embodiment of the present application does not limit the number of field mapping models, and can be set according to the number of signal types in actual business; in order to achieve targeted and accurate mapping of vehicle signals of corresponding signal types, each field mapping model may adopt a different network structure.

[0030] Exemplarily, taking the example that the signal type includes a numerical type and a text type, a field mapping model can be pre-trained for a vehicle signal of the numerical type, and another field mapping model can be pre-trained for a vehicle signal of the text type.

[0031] In a specific implementation, the multiple field mapping models include a convolutional neural network model and a pre-trained language representation model, and determining a target model from the multiple trained field mapping models based on the signal type includes: Determine whether the signal type is a text type; if so, determine the trained pre-trained language representation model as the target model; otherwise, determine the trained convolutional neural network model as the target model.

[0032] The convolutional neural network model may be a CNN (Convolutional Neural Network) model, and the pre-trained language representation model may be a BERT (Bidirectional Encoder Representations from Transformers) model.

[0033] Specifically, for vehicle signals of numerical type, a convolutional neural network model can be pre-trained as the corresponding field mapping model to process the complex numerical mapping relationship of vehicle signals of numerical type; for vehicle signals of text type, a pre-trained language representation model can be pre-trained as the corresponding field mapping model to process the semantic content mapping relationship of vehicle signals of text type.

[0034] In an embodiment of the present application, after determining the signal type corresponding to the current vehicle signal, it is possible to first determine whether the signal type is a text type. If so, the pre-trained language representation model is determined as the target model; if not, the convolutional neural network model is determined as the target model.

[0035] In this embodiment, by selecting a convolutional neural network model to identify the field information of a numerical type of vehicle signal, the complex numerical mapping relationship between signals can be processed, and by selecting a pre-trained language representation model to identify the field information of a text type of vehicle signal, the semantic content of the vehicle signal can be more effectively processed and understood, thereby improving the accuracy of vehicle signal field mapping.

[0036] It should be noted that for text types, the corresponding field mapping model is not limited to the above-mentioned pre-trained language representation model, but can also be other models that can process semantic features, such as Recurrent Neural Network (RNN) language model, Long Short-Term Memory (LSTM) model, Transformer model, Generative Pretrained Transformer (GPT) model, etc.

[0037] Moreover, for numeric types, the corresponding field mapping model is not limited to the above-mentioned convolutional neural network model, but can also be other models that can process waveform features, such as generative adversarial networks (GAN) models, deep belief networks (DBN) models, autoencoder (AE) models, etc.

[0038] In one example, the training process of the field mapping model includes the following steps: Step 11: Collect sample vehicle signals of each vehicle type and determine key metadata of each sample vehicle signal; Step 12: for each sample vehicle signal of the numerical type, determine the waveform feature of the sample vehicle signal based on the key metadata of the sample vehicle signal, and for each sample vehicle signal of the text type, determine the semantic feature of the sample vehicle signal based on the key metadata of the sample vehicle signal; Step 13: Based on the waveform features of each sample vehicle signal of the numerical type and the corresponding field labels, the convolutional neural network is trained, and based on the semantic features of each sample vehicle signal of the text type and the corresponding field labels, the pre-trained language representation network is trained to obtain a convolutional neural network model and a pre-trained language representation model.

[0039] Among them, in step 11, real-time vehicle signals can be captured from vehicle control systems of various models as sample vehicle signals through an integrated multi-protocol interface and a streaming data processing method. After acquiring each sample vehicle signal, an adaptive algorithm can be used to perform data cleaning, denoising and standardization on all sample vehicle signals to ensure signal quality and consistency, providing a reliable basis for subsequent model training.

[0040] Furthermore, for each sample vehicle signal, key metadata of the sample vehicle signal may be extracted, wherein the key metadata may include a signal name, a signal identifier (ID), a data type, a physical range, a unit, and a signal value content.

[0041] It should be noted that the purpose of extracting key metadata of sample vehicle signals is to analyze the correlation between different vehicle signals and the value characteristics of vehicle signals through key metadata. By extracting key metadata, it is possible to avoid analyzing all metadata of the signal, thereby reducing the data dimension while ensuring the accuracy of subsequent model training.

[0042] Furthermore, in step 12, for the sample vehicle signals of numerical type, corresponding waveform features can be extracted through key metadata, wherein the waveform features include amplitude features, frequency features and phase features. Also, for the sample vehicle signals of text type, corresponding semantic features can be extracted through key metadata, such as mining semantic features from information such as signal name, signal description, signal value content, etc. in the key metadata.

[0043] After obtaining the characteristics of the sample vehicle signals, dimensionality reduction methods such as principal component analysis and linear discriminant analysis can be used to process the characteristics of all sample vehicle signals to achieve feature dimensionality reduction and optimize feature expression. In addition, mutual information, chi-square test and other methods can be used for feature selection to improve the efficiency of subsequent model training.

[0044] Furthermore, in step 13, the field mapping model can be trained according to the characteristics of each sample vehicle signal and the corresponding field label. The field label corresponding to the sample vehicle signal is the real field content of the sample vehicle signal, and the field label can be determined by domain expert knowledge, which can cover the signal correspondence between a wide range of vehicle models. Alternatively, the field label corresponding to the sample vehicle signal can also be obtained by manual annotation, and the mapping relationship between signal fields of different vehicle models can be clarified by manual annotation.

[0045] Specifically, the convolutional neural network can be trained according to the waveform characteristics of each sample vehicle signal of the numerical type and the corresponding field labels to obtain a convolutional neural network model, and the pre-trained language representation network can be trained according to the semantic characteristics of each sample vehicle signal of the text type and the corresponding field labels to obtain a pre-trained language representation model.

[0046] During the training process, all sample vehicle signals can be divided according to the proportion, such as 70% as the training set, 15% as the validation set, and the remaining 15% as the test set. First, the training set is used to train the convolutional neural network model and the pre-trained language representation model. For example, the characteristics of the sample vehicle signal are input into the network, and the loss value between the prediction result output by the network and the field label is calculated. Based on the loss value, the network parameters are tuned by the back propagation algorithm until the loss value converges; further, the validation set can be used to verify the prediction accuracy of the convolutional neural network model and the pre-trained language representation model; finally, the test set is used to optimize the convolutional neural network model and the pre-trained language representation model. For example, the model is optimized by using cross-validation, hyperparameter tuning and other methods to improve the generalization ability and stability of the model and prevent overfitting.

[0047] Based on the above steps 11 to 13, a convolutional neural network model and a pre-trained language representation model can be trained. By extracting the key metadata of the signal, the waveform features or semantic features of the signal are extracted, and a convolutional neural network model for predicting numerical type vehicle signals and a pre-trained language representation model for predicting text type vehicle signals are trained respectively. While improving the model training efficiency, the field mapping reliability of the field mapping model can also be guaranteed.

[0048] In the embodiment of the present application, after the model is trained and deployed, considering that as car companies and car manufacturers develop new models and optimize old models, vehicle signals of new models may be generated or vehicle signals of old models may change, after the convolutional neural network model and the pre-trained language representation model are trained, new data sets can be continuously introduced to optimize the model for training, so as to improve the mapping accuracy and generalization ability of the model.

[0049] Based on the above implementations, optionally, after obtaining the convolutional neural network model and the pre-trained language representation model, the following steps are further included: Step 14: Obtain optimized training data set; Step 15: Based on the vehicle signals of numerical type and the corresponding field labels in the optimized training data set, the convolutional neural network model is optimized and trained to update the convolutional neural network model; Step 16: Based on the vehicle signals of text type and the corresponding field labels in the optimized training data set, the pre-trained language representation model is optimized and trained to update the pre-trained language representation model.

[0050] The optimized training data set may include vehicle signals of new models and vehicle signals of changed old models. In addition, the optimized training data set also includes field labels corresponding to each vehicle signal.

[0051] Specifically, new models and old models with changed vehicle signals can be determined in advance, and then all vehicle signals of new models and changed vehicle signals of old models are collected, and the optimized training data set is constructed in combination with the corresponding field labels. In step 14, the optimized training data set can be directly obtained.

[0052] In step 15 and step 16, the convolutional neural network model can be further trained by optimizing the vehicle signals of numerical type and the corresponding field labels in the training data set to tune the model parameters and achieve the purpose of updating the convolutional neural network model; and the pre-trained language representation model can be further trained by optimizing the vehicle signals of text type and the corresponding field labels in the training data set to tune the model parameters and achieve the purpose of updating the pre-trained language representation model.

[0053] Through the above steps 14 to 16, the convolutional neural network model and the pre-trained language representation model can be retrained by continuously introducing new labeled data to improve the mapping accuracy and generalization ability of the model. In addition, by updating the model online, it can be ensured that the model can adapt to changes in vehicle signals and updates to mapping requirements in real time.

[0054] In an embodiment of the present application, the user can also actively send vehicle model change information when the vehicle signal of certain vehicle models changes, and update at least one of the convolutional neural network model and the pre-trained language representation model.

[0055] On the basis of the above implementation modes, optionally, the method provided in the embodiment of the present application further includes: In response to detecting the vehicle model change information, a vehicle signal of the changed vehicle model corresponding to the vehicle model change information is acquired; and based on the vehicle signal of the changed vehicle model and the corresponding field information, at least one field mapping model is updated.

[0056] Among them, the vehicle model change information may include the identification of the vehicle model whose vehicle signal has changed, such as the change in the signal name of the vehicle signal, the change in the signal description of the vehicle signal, the change in the signal value content of the vehicle signal (such as the change in the value range of the numerical value, the change in the value type of the numerical value, the change in the word limit of the text, etc.), the change in the field information corresponding to the vehicle signal, etc.

[0057] Specifically, users can actively upload vehicle model change information in the system. If vehicle model change information is detected, the changed vehicle model corresponding to the vehicle model change information can be determined, and then the vehicle signal of the changed vehicle model can be collected. The vehicle signal of the changed vehicle model is classified according to the signal type. If there is a numerical type vehicle signal, the convolutional neural network model is updated; if there is a text type vehicle signal, the pre-trained language representation model is updated.

[0058] Through the above optional implementation, the field mapping model can be optimized based on the vehicle model change information actively sent by the user, ensuring that the field mapping model can adapt to changes in vehicle signals and updates to mapping requirements, further improving the mapping accuracy of the model.

[0059] In an embodiment of the present application, for the current vehicle signal, after selecting a target model based on the signal type corresponding to the current vehicle signal, further, the current vehicle signal can be subjected to field recognition according to the target model, i.e., field mapping, to determine the field information corresponding to the current vehicle signal. The field information can specifically be the field content corresponding to the signal value content in the current vehicle signal, such as vehicle speed, acceleration, gear position, etc. The field information can include a field content, such as a vehicle signal that describes a field alone or multiple vehicle signals that describe a field together; or, the field information can include multiple field contents, such as a vehicle signal that describes multiple fields alone.

[0060] Specifically, determining the field information corresponding to the current vehicle signal based on the target model can be: performing data cleaning, denoising and standardization processing on the current vehicle signal, and then extracting the key metadata of the current vehicle signal, determining the waveform features or semantic features of the current vehicle signal through the key metadata of the current vehicle signal, and then inputting the waveform features or semantic features into the target model to obtain the field information output by the target model.

[0061] In an embodiment of the present application, taking into account the possibility of incorrect model mapping, in order to further optimize the model, after determining the field information corresponding to the current vehicle signal based on the target model, the field information can be displayed for user confirmation. If the user confirms that it is incorrect, such signal can be used to continue optimizing the model.

[0062] Based on the above implementations, optionally, after determining the field information corresponding to the current vehicle signal based on the target model, the following steps are further included: Step 21: Display the current vehicle signal and corresponding field information in the signal display interface; Step 22, in response to detecting that the user triggers a confirmation abnormality operation for the current vehicle signal in the signal display interface, the current vehicle signal is determined as a negative sample signal, and the corresponding field information is determined as the label corresponding to the negative sample signal, so as to update the target model based on the negative sample information and the corresponding label.

[0063] Specifically, in step 21, after the field information corresponding to the current vehicle signal is identified by the target model, the current vehicle signal and the corresponding field information can be displayed through the signal display interface for user confirmation. The signal display interface can be an interface for users to browse the field mapping results of each vehicle signal.

[0064] Furthermore, in step 22, if it is detected that the user triggers an abnormal confirmation operation for the current vehicle signal in the signal display interface, it means that the user confirms that the field information corresponding to the current vehicle signal is incorrect, that is, the model mapping is incorrect. At this time, the current vehicle signal can be determined as a negative sample signal, and the corresponding field information can be determined as the label corresponding to the negative sample signal. Subsequently, the negative sample information and the corresponding label can be used to distinguish the signal type to optimize the training target model.

[0065] In the above steps 21 and 22, after the fields are mapped through the target model, the current vehicle signal and the corresponding field information can be displayed for user confirmation, and when the user feedback is abnormal, it can be promptly confirmed as a negative sample to facilitate subsequent optimization of the model, thereby further ensuring the mapping accuracy of the model.

[0066] In addition to displaying the current vehicle signal and the corresponding field information and optimizing the model according to the anomalies reported by the user, it can also automatically detect whether the field mapping of the target model is incorrect.

[0067] In a specific implementation, after determining the field information corresponding to the current vehicle signal based on the target model, the following steps are also included: Step 31, detecting whether the field information exists in the preset field table, and whether the signal value content of the current vehicle signal satisfies the content rule corresponding to the field information in the preset field table; Step 32: In response to the field information not existing in the preset field table, or the signal value content not satisfying the content rule, abnormal prompt information is displayed on the signal display interface, and an abnormal prompt signal is sent to the associated terminal device.

[0068] Among them, in step 31, the preset field table may include pre-defined field information and content rules corresponding to each field information; the content rules may specify the size, value range and number of words of the signal value content of the vehicle signal describing the field information.

[0069] Specifically, after identifying the field information corresponding to the current vehicle signal through the target model, it can be detected whether the field information exists in the preset field table. If so, it is further determined whether the signal value content of the current vehicle signal satisfies the content rules corresponding to the field information in the table, such as determining whether the size of the signal value content satisfies the size specified in the content rule, or determining whether the value of the signal value content satisfies the value range specified in the content rule, etc.

[0070] Furthermore, in step 32, if it is detected that the field information does not exist in the preset field table, it can be confirmed that the target model has identified an undefined field; if it is detected that the signal value content does not meet the corresponding content rules, such as the number of text words exceeds the limit, the value is not within the range, etc., it can be confirmed that the field information recognized by the target model does not conform to its corresponding content rules, and the target model recognizes the field incorrectly.

[0071] When the target model identifies an undefined field or an incorrectly identified field, an abnormal prompt message can be displayed through the signal display interface to remind the user to handle it, wherein the abnormal prompt message can describe the current vehicle signal and the corresponding field content. In addition to displaying the abnormal prompt message, an abnormal prompt signal can also be sent to an associated terminal device to remind the user to handle it, wherein the associated device can be a user terminal.

[0072] In addition to displaying abnormal prompt information and sending abnormal prompt signals on the display interface, the current vehicle signal and the corresponding field content can also be recorded, such as stored in an abnormal signal record table, and the abnormal signal record table can be used to optimize the training field mapping model later.

[0073] Through the above steps 31 and 32, it is possible to detect situations where the target model identifies undefined fields and where the target model identifies incorrect fields. By displaying abnormal prompt information and sending abnormal prompt signals, the user can be reminded to handle the situation in time to avoid abnormal field mapping results affecting subsequent data analysis.

[0074] In an embodiment of the present application, after obtaining the field information corresponding to the current vehicle signal based on the target model, the current vehicle signal can also be converted into a signal in a standardized format through the field information, and then the converted signal can be stored in a standardized database to output a standardized signal format, thereby achieving seamless docking and data sharing between different systems, and promoting the interoperability and utilization of vehicle signals between multiple systems and platforms.

[0075] Considering that in addition to the cases where the model recognizes undefined fields and the model recognizes incorrectly, it is also possible that the model has never recognized a pre-defined field. Therefore, the standardized database can also be tested to detect whether there is a case where the model does not recognize a field, so as to prompt the user.

[0076] In another specific implementation, after determining the field information corresponding to the current vehicle signal based on the target model, the following steps are also included: Step 41: Convert the current vehicle signal into a format based on the field information to obtain a target signal, and store the target signal in association with the corresponding field information in the standardized database; Step 42: Check whether there is a field in the standardized database that has not been associated with the target signal. If so, display abnormal prompt information on the signal display interface and send an abnormal prompt signal to the associated terminal device.

[0077] Among them, in step 41, the format of the current vehicle signal can be converted according to the field information, and the converted signal is stored as a target signal in association with the corresponding field information in the standardized database.

[0078] Furthermore, it is possible to detect whether there are fields in the standardized database that have not been associated with the target signal. If so, it means that the field may have never been identified due to incorrect model identification. At this time, abnormal prompt information can be displayed through the signal display interface. The abnormal prompt information may include fields that have not been associated with the target signal, and an abnormal prompt signal can also be sent to the associated terminal device.

[0079] In addition to displaying abnormal prompt information and sending abnormal prompt signals on the display interface, you can also record the fields that have not been associated with the target signal, and then continue to optimize the training field mapping model based on this.

[0080] Through the above steps 41 and 42, the vehicle signal can be formatted and output in a standardized signal format can be achieved, which is convenient for subsequent seamless connection and data sharing between different systems. In addition, by detecting whether there are fields in the standardized database that have not been associated with the target signal, it can be determined whether there are unrecognized fields, so that the user can be reminded to process in time to avoid affecting subsequent data analysis.

[0081] The field mapping method of vehicle signals provided in the embodiment of the present application obtains the current vehicle signal of the vehicle model to be mapped, and determines the signal type corresponding to the current vehicle signal, and then determines the target model in the multiple field mapping models that have been trained according to the signal type, and determines the field information corresponding to the current vehicle signal through the target model, thereby realizing model-based signal field mapping, eliminating the need for manual comparison of vehicle signal matrices, greatly improving the mapping efficiency and accuracy of vehicle signals, thereby improving the efficiency and accuracy of large-scale vehicle signal integration and analysis, and improving the processing efficiency of vehicle signals, ensuring the real-time processing of vehicle signals, and reducing delays. In addition, the method takes into account the different signal characteristics of different signal types, and identifies the field information through the field mapping model corresponding to the signal type, which can further improve the field mapping accuracy.

[0082] The present application also provides a field mapping system for vehicle signals, such as Figure 2 As shown, Figure 2 It is a schematic diagram of a field mapping system for vehicle signals provided in an embodiment of the present application.

[0083] Among them, the field mapping system of vehicle signals includes data acquisition and intelligent preprocessing module, feature engineering and intelligent selection module, model and adaptive training module, and real-time mapping and dynamic adjustment module.

[0084] Specifically, for the data acquisition and intelligent preprocessing module, it can integrate multi-protocol interfaces (such as CAN, LIN, Ethernet, etc.), automatically capture real-time vehicle signals from vehicle control systems of various models, and use Flink streaming data processing to ensure that the signals are captured and processed at the same time they are generated, reducing delays and improving system response speed; then the signals are adaptively cleaned, including outlier and missing value processing, similar sample deduplication, etc., and metadata extraction is performed, and finally the cleaned signals and the extracted key metadata are output to the feature engineering and intelligent selection module.

[0085] For the feature engineering and intelligent selection module, after receiving the data, it can perform feature extraction, such as analyzing the waveform characteristics (amplitude, frequency, phase, etc.) of numerical type signals and analyzing the text characteristics of text type signals; then perform feature selection and dimensionality reduction, specifically using principal component analysis, linear discriminant analysis and other dimensionality reduction techniques; finally, output the feature vector to the model and adaptive training module.

[0086] For the model and adaptive training module, after receiving the feature vector, it can first build a network (CNN and BERT), then label the signal, build a training set, and perform model training and optimization, including cross-validation and hyperparameter tuning. The trained model can be deployed to the system, and the model can also be continuously learned and optimized.

[0087] For the feature engineering and intelligent selection module, after receiving the vehicle signal of the vehicle model to be mapped, it can use the model to identify the field information of the signal, and can also adaptively adjust the model according to the identification result, such as dynamically adjusting the mapping rules. After the model identifies the field information, it can convert the signal format and output it in a standardized database.

[0088] The field mapping system of the above-mentioned vehicle signal has the following technical effects: 1. Improved signal processing efficiency and real-time performance: integrated multi-protocol interface and streaming data processing technology to achieve comprehensive and real-time capture and processing, reduce delays, and improve system response speed; 2. Improved signal quality and consistency: adaptive algorithms are used for cleaning, denoising and standardization to ensure high signal quality and consistency, providing a reliable basis for subsequent analysis; 3. Enhanced the mapping accuracy and generalization ability of the model: through feature engineering and intelligent selection technology, the metadata characteristics of the signal are deeply mined, and the model is combined to realize automatic matching and mapping of signal fields, improving mapping accuracy and generalization ability; 4. Continuous optimization and adaptation of mapping rules are realized: the introduction of adaptive learning and continuous optimization mechanisms enables the system to adapt to new vehicle models and signal changes, maintain the timeliness and accuracy of mapping results, and reduce maintenance costs; 5. Promoted data sharing and interoperability: output standardized signal formats, realize seamless docking and signal sharing between different systems, and promote the interoperability and utilization of vehicle signals between multiple systems and platforms.

[0089] Figure 3 Schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 3 As shown, the electronic device 400 includes one or more processors 401 and a memory 402 .

[0090] The processor 401 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 400 to perform desired functions.

[0091] The memory 402 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 401 may run the program instructions to implement the field mapping method of the vehicle signal of any embodiment of the present application described above and / or other desired functions. Various contents such as initial external parameters, thresholds, etc. may also be stored in the computer-readable storage medium.

[0092] In one example, the electronic device 400 may further include: an input device 403 and an output device 404, which are interconnected via a bus system and / or other forms of connection mechanisms (not shown). The input device 403 may include, for example, a keyboard, a mouse, etc. The output device 404 may output various information to the outside, including early warning prompt information, braking force, etc. The output device 404 may include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, etc.

[0093] Of course, to simplify, Figure 3 Only some of the components related to the present application in the electronic device 400 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application conditions, the electronic device 400 may also include any other appropriate components.

[0094] In addition to the above-mentioned methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the field mapping method of the vehicle signal provided in any embodiment of the present application.

[0095] The computer program product may be written in any combination of one or more programming languages ​​to write program codes for performing the operations of the embodiments of the present application, including object-oriented programming languages, such as Java, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0096] In addition, an embodiment of the present application may also be a computer-readable storage medium having computer program instructions stored thereon, and when the computer program instructions are executed by a processor, the processor executes the steps of the field mapping method of the vehicle signal provided in any embodiment of the present application.

[0097] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable 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.

[0098] It should be noted that the terms used in this application are only for describing specific embodiments, rather than limiting the scope of this application. As shown in the specification and claims of this application, unless the context clearly indicates an exception, the words "one", "a", "a kind of" and / or "the" do not specifically refer to the singular, but may also include the plural. The terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the presence of other identical elements in the process, method or device including the elements.

[0099] It should also be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present application. Unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", etc. should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0100] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and its core ideas of this application. The above is only the preferred implementation method of this application. It should be pointed out that due to the limitations of textual expression and the objective existence of infinite specific structures, ordinary technicians in this technical field can make several improvements, modifications or changes without departing from the principles of this application, and can also combine the above technical features in an appropriate manner; these improvements, modifications, changes or combinations, or the direct application of the concept and technical solution of the invention to other occasions without improvement, should be regarded as the scope of protection of this application.

Claims

1. A field mapping method for vehicle signals, characterized in that: include: Obtaining a current vehicle signal of a vehicle model to be mapped, and determining a signal type corresponding to the current vehicle signal; A target model is determined from among multiple trained field mapping models based on the signal type, and field information corresponding to the current vehicle signal is determined based on the target model.

2. The method according to claim 1, characterized in that The multiple field mapping models include a convolutional neural network model and a pre-trained language representation model, and determining a target model from the multiple trained field mapping models based on the signal type includes: Determining whether the signal type is a text type; If so, the trained pre-trained language representation model is determined as the target model; otherwise, the trained convolutional neural network model is determined as the target model.

3. The method according to claim 2, characterized in that The training process of the field mapping model includes: Collect sample vehicle signals of each vehicle type and determine key metadata of each sample vehicle signal; For each sample vehicle signal of a numerical type, determining a waveform feature of the sample vehicle signal based on key metadata of the sample vehicle signal, and for each sample vehicle signal of a text type, determining a semantic feature of the sample vehicle signal based on the key metadata of the sample vehicle signal; Based on the waveform features of each sample vehicle signal of numerical type and the corresponding field labels, a convolutional neural network is trained, and based on the semantic features of each sample vehicle signal of text type and the corresponding field labels, a pre-trained language representation network is trained to obtain the convolutional neural network model and the pre-trained language representation model.

4. The method according to claim 3, characterized in that After obtaining the convolutional neural network model and the pre-trained language representation model, the method further includes: Obtain optimized training data set; Based on the vehicle signals of numerical type and the corresponding field labels in the optimized training data set, the convolutional neural network model is optimized and trained to update the convolutional neural network model; Based on the text-type vehicle signals and corresponding field labels in the optimized training data set, the pre-trained language representation model is optimized and trained to update the pre-trained language representation model.

5. The method according to claim 1, characterized in that The method further comprises: In response to detecting the vehicle model change information, acquiring a vehicle signal of a changed vehicle model corresponding to the vehicle model change information; At least one of the field mapping models is updated based on the vehicle signal of the changed vehicle model and the corresponding field information.

6. The method according to claim 1, characterized in that After determining the field information corresponding to the current vehicle signal based on the target model, the method further includes: Displaying the current vehicle signal and corresponding field information in the signal display interface; In response to detecting that a user triggers a confirmation abnormality operation for the current vehicle signal in the signal display interface, the current vehicle signal is determined as a negative sample signal, and the corresponding field information is determined as a label corresponding to the negative sample signal, so as to update the target model based on the negative sample information and the corresponding label.

7. The method according to claim 1, characterized in that After determining the field information corresponding to the current vehicle signal based on the target model, the method further includes: Detecting whether the field information exists in a preset field table, and whether the signal value content of the current vehicle signal satisfies a content rule corresponding to the field information in the preset field table; In response to the field information not existing in the preset field table, or the signal value content not satisfying the content rule, abnormal prompt information is displayed on the signal display interface, and an abnormal prompt signal is sent to the associated terminal device.

8. The method according to claim 1, characterized in that After determining the field information corresponding to the current vehicle signal based on the target model, the method further includes: Performing format conversion on the current vehicle signal based on the field information to obtain a target signal, and associating and storing the target signal under the corresponding field information in a standardized database; Check whether there is a field in the standardized database that has not been associated with the target signal. If so, display abnormal prompt information on the signal display interface and send an abnormal prompt signal to the associated terminal device.

9. An electronic device, characterized in that: The electronic device comprises: Processor and memory; The processor is used to execute the steps of the field mapping method of the vehicle signal as described in any one of claims 1 to 8 by calling the program or instruction stored in the memory.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program or an instruction, and the program or the instruction enables a computer to execute the steps of the field mapping method of a vehicle signal as claimed in any one of claims 1 to 8.

Images