Vehicle power supply equipment prediction method and system, vehicle networking control equipment and vehicle

Abstract

The invention provides a vehicle power supply equipment prediction method and system, vehicle networking control equipment and a vehicle. The vehicle power supply equipment prediction method comprises the steps that charging data and power supply equipment data of a target vehicle are acquired; wherein the charging data comprises charging power and charging time, the power supply equipment data comprises the position of power supply equipment and the power of the power supply equipment, and the power supply equipment is used for supplying power to the target vehicle; a model framework is constructed based on a variational auto-encoder and a self-attention model, and training is performed by using the charging data, the power supply equipment data and the model framework to generate a sequence prediction model; and finally, predicting the operation state of the power supply equipment in the target area by using the sequence prediction model. According to the method, the running state of the vehicle power supply equipment is predicted by constructing the sequence prediction model, the limitation of a traditional method in a complex charging scene is effectively solved, the prediction accuracy and stability are improved, and more effective support is provided for management and scheduling of the vehicle power supply equipment.

Description

Technical Field

[0001] The present invention relates to the field of vehicles and power supply technology, and in particular to a vehicle power supply equipment prediction method, system, vehicle networking control equipment and vehicle. Background Art

[0002] Traditional prediction methods for electric vehicle power supply equipment are mainly based on rules or statistical analysis, such as simple regression models based on historical data or moving average methods based on time series. Although these methods can predict the use of charging equipment to a certain extent, they face several obvious shortcomings. First, traditional methods can only consider a single variable or simple rules, and cannot effectively capture the complex dynamic changes and the correlation between multiple variables in the charging scene. Secondly, traditional methods have limited use of real-time information and cannot adjust the prediction model in time to adapt to changes in actual scenarios. In addition, traditional methods have poor robustness in dealing with noisy data and abnormal situations, and are easily affected by data quality and collection methods, resulting in unstable or inaccurate prediction results. The deployment and update efficiency are low. The deployment and online update of traditional models are usually cumbersome and difficult to meet the needs of real-time prediction. Summary of the invention

[0003] In view of the above-mentioned shortcomings of the prior art, an object of the present invention is to provide a vehicle power supply equipment prediction method, system, vehicle networking control device and vehicle, so as to solve the technical problems existing in the prior art.

[0004] To achieve the above-mentioned object and other related objects, the present invention provides a vehicle power supply equipment prediction method, comprising the following steps:

[0005] Obtain charging data and power supply equipment data of the target vehicle; wherein the charging data includes charging power and charging time, and the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle;

[0006] Building a model framework based on a variational autoencoder and a self-attention model, and using the charging data, the power supply device data, and the model framework for training to generate a sequence prediction model;

[0007] The sequence prediction model is used to predict the operating status of the power supply equipment in the target area.

[0008] In one embodiment of the present invention, the process of generating a sequence prediction model includes:

[0009] A dual-stream variational autoencoder framework is constructed based on the variational autoencoder, and the charging data is input into a first variational autoencoder in the dual-stream variational autoencoder framework, and the power supply device data is input into a second variational autoencoder in the dual-stream variational autoencoder framework;

[0010] A coding and decoding structure unit is established based on the encoder and decoder in the self-attention model, and the processing result of the first variational autoencoder is output to the first encoder, and the processing result of the second variational autoencoder is output to the second encoder; and the first encoder and the second encoder are used to extract time series multi-scale information features, and the time series multi-scale information features extracted based on the first encoder are input into the first decoder for time series feature fusion, and the time series multi-scale information features extracted based on the second encoder are input into the second decoder for time series feature fusion;

[0011] Establishing an interactive learning structure unit, and inputting the processing result of the first variational autoencoder, the processing result of the second variational autoencoder, the time series multi-scale information features extracted by the first encoder, and the time series multi-scale information features extracted by the second encoder into the interactive learning structure unit for interactive learning training to obtain an interactive learning training result;

[0012] Inputting the time series multi-scale information features from the first encoder and the second encoder and the interactive learning training results into the third decoder simultaneously for time series feature fusion;

[0013] The sequence prediction model is generated according to the time series feature fusion result of the first decoder, the time series feature fusion result of the second decoder, and the time series feature fusion result of the third decoder.

[0014] In one embodiment of the present invention, the process of using the sequence prediction model to predict the operating status of the power supply equipment in the target area includes: storing the sequence prediction model as a model file, and deploying it based on the model file to predict the operating status of the power supply equipment in the target area; wherein the deployment environment of the model file includes an application.

[0015] In one embodiment of the present invention, the process of deploying based on the model file and predicting the operating status of the power supply equipment in the target area includes:

[0016] Converting the model file into a preset file format to obtain a preset file format model;

[0017] Obtain a folder created in advance or in real time for loading the preset file format model, and install dependencies in the folder;

[0018] Using a script created in advance or in real time as an inference prediction entry for calling the preset file format model, the script is used to process input data from an external access request and perform model inference prediction on a local end;

[0019] An application program interface service is created based on the dependencies, and an access request fed back by a front-end page is received, the script is called to load the preset file format model, the operating status of the power supply equipment in the target area is predicted, and the power supply equipment prediction result is fed back to the front-end page for display.

[0020] In one embodiment of the present invention, the process of acquiring charging data and power supply equipment data of a target vehicle includes:

[0021] Select the collection equipment according to the data collection type and data collection target determined in advance or in real time;

[0022] Based on a collection frequency, collection method and data format determined in advance or in real time, the charging data of the target vehicle and the power supply device data are collected by the collection device;

[0023] The charging data and power supply equipment data collected by the collection device are cleaned, and the pre-processed charging data and power supply equipment data at the current timestamp are aggregated according to the time aggregation scale, and the occupancy rate of the power supply equipment is calculated, and the data is randomly segmented; the data cleaning includes: removing duplicate data, filling missing values, and correcting erroneous data.

[0024] In one embodiment of the present invention, the target vehicle includes a new energy vehicle, and the power supply equipment includes a charging pile.

[0025] The present invention also provides a vehicle power supply equipment prediction system, the system comprising:

[0026] A data acquisition module, used to obtain charging data and power supply equipment data of the target vehicle; wherein the charging data includes charging power and charging time, and the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle;

[0027] A model training module, used to build a model framework based on a variational autoencoder and a self-attention model, and to train using the charging data, the power supply device data, and the model framework to generate a sequence prediction model;

[0028] The prediction module is used to predict the operating status of the power supply equipment in the target area by using the sequence prediction model.

[0029] In one embodiment of the present invention, the process of generating a sequence prediction model by the model training module includes:

[0030] A dual-stream variational autoencoder framework is constructed based on the variational autoencoder, and the charging data is input into a first variational autoencoder in the dual-stream variational autoencoder framework, and the power supply device data is input into a second variational autoencoder in the dual-stream variational autoencoder framework;

[0031] A coding and decoding structure unit is established based on the encoder and decoder in the self-attention model, and the processing result of the first variational autoencoder is output to the first encoder, and the processing result of the second variational autoencoder is output to the second encoder; and the first encoder and the second encoder are used to extract time series multi-scale information features, and the time series multi-scale information features extracted based on the first encoder are input into the first decoder for time series feature fusion, and the time series multi-scale information features extracted based on the second encoder are input into the second decoder for time series feature fusion;

[0032] Establishing an interactive learning structure unit, and inputting the processing result of the first variational autoencoder, the processing result of the second variational autoencoder, the time series multi-scale information features extracted by the first encoder, and the time series multi-scale information features extracted by the second encoder into the interactive learning structure unit for interactive learning training to obtain an interactive learning training result;

[0033] Inputting the time series multi-scale information features from the first encoder and the second encoder and the interactive learning training results into the third decoder simultaneously for time series feature fusion;

[0034] The sequence prediction model is generated according to the time series feature fusion result of the first decoder, the time series feature fusion result of the second decoder, and the time series feature fusion result of the third decoder.

[0035] The present invention also provides a vehicle networking control device, including a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of any one of the above-mentioned vehicle power supply equipment prediction methods.

[0036] The present invention also provides a vehicle, characterized in that the vehicle includes the vehicle power supply equipment prediction system as described above, or includes the vehicle networking control device as described above.

[0037] As described above, the present invention provides a vehicle power supply equipment prediction method, system, vehicle networking control device and vehicle, which have the following beneficial effects: the present invention first obtains the charging data and power supply equipment data of the target vehicle; wherein the charging data includes the charging power and the charging time, and the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle; then, a model framework is constructed based on a variational autoencoder and a self-attention model, and the charging data, the power supply equipment data and the model framework are used for training to generate a sequence prediction model; finally, the sequence prediction model is used to predict the operating status of the power supply equipment in the target area. It can be seen that the present invention predicts the operating status of the vehicle power supply equipment by constructing a sequence prediction model, which not only effectively solves the limitations of traditional methods in the face of complex charging scenarios, but also improves the accuracy and stability of the prediction. In addition, the present invention can realize the real-time deployment of the model and the dynamic update of the prediction results, providing more effective support for the management and scheduling of vehicle power supply equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A schematic diagram of an exemplary system architecture for applying the technical solutions in one or more embodiments of the present invention;

[0039] Figure 2 A schematic diagram of a flow chart of a vehicle power supply equipment prediction method provided in an embodiment of the present invention;

[0040] Figure 3 A schematic flow chart of a vehicle power supply equipment prediction method provided in another embodiment of the present invention;

[0041] Figure 4 A schematic diagram of the framework structure of a sequence prediction model provided in one embodiment of the present invention;

[0042] Figure 5 A schematic diagram of an electric vehicle power supply equipment prediction application process provided by an embodiment of the present invention;

[0043] Figure 6 A visualization application diagram showing prediction of electric vehicle power supply equipment provided by an embodiment of the present invention;

[0044] Figure 7 A schematic diagram of the hardware structure of a vehicle power supply equipment prediction system provided by an embodiment of the present invention;

[0045] Figure 8 The figure is a schematic diagram of the hardware structure of a vehicle networking control device suitable for implementing one or more embodiments of the present invention. DETAILED DESCRIPTION

[0046] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0047] It should be noted that the illustrations provided in this embodiment are only used to illustrate the basic concept of the present invention in a schematic manner. Therefore, the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0048] Figure 1 FIG. 1 is a schematic diagram showing an exemplary system architecture to which the technical solution in one or more embodiments of the present invention can be applied. Figure 1 As shown, the system architecture 100 may include a terminal device 110, a network 120, and a server 130. The terminal device 110 may include various electronic devices such as a smart phone, a tablet computer, a laptop computer, and a desktop computer. The server 130 may be an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The network 120 may be a communication medium of various connection types that can provide a communication link between the terminal device 110 and the server 130, such as a wired communication link or a wireless communication link.

[0049] According to the implementation requirements, the system architecture in the embodiment of the present invention may have any number of terminal devices, networks and servers. For example, the server 130 may be a server group composed of multiple server devices. In addition, the technical solution provided in the embodiment of the present invention may be applied to the terminal device 110, or to the server 130, or may be implemented by the terminal device 110 and the server 130 together, and the present invention does not make any special limitation on this.

[0050] In one embodiment of the present invention, the terminal device 110 or server 130 of the present invention can obtain the charging data and power supply equipment data of the target vehicle; wherein the charging data includes the charging power and the charging time, and the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle; then a model framework is constructed based on the variational autoencoder and the self-attention model, and the charging data, the power supply equipment data and the model framework are used for training to generate a sequence prediction model; finally, the sequence prediction model is used to predict the operating status of the power supply equipment in the target area. By using the terminal device 110 or the server 130 to execute the vehicle power supply equipment prediction method, the operating status of the vehicle power supply equipment can be predicted by constructing a sequence prediction model, which not only effectively solves the limitations of traditional methods when facing complex charging scenarios, but also improves the accuracy and stability of the prediction. In addition, the present invention can realize the real-time deployment of the model and the dynamic update of the prediction results, providing more effective support for the management and scheduling of vehicle power supply equipment.

[0051] The above section introduces the contents of an exemplary system architecture applying the technical solution of the present invention. Next, the vehicle power supply equipment prediction method of the present invention will be introduced.

[0052] Figure 2 FIG. 1 is a flow chart of a method for predicting a vehicle power supply device according to an embodiment of the present invention. Specifically, in an exemplary embodiment, Figure 2 As shown, this embodiment provides a vehicle power supply equipment prediction method, the method comprising the following steps:

[0053] S210, acquiring charging data and power supply equipment data of the target vehicle; wherein the charging data includes charging power and charging time, and the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle. As an example, in this embodiment or other embodiments, the target vehicle includes but is not limited to a new energy vehicle, and the power supply equipment includes but is not limited to a charging pile. For example, in this embodiment, the new energy vehicle may be an electric vehicle.

[0054] S220, construct a model framework based on the variational autoencoder and the self-attention model, and train using the charging data, the power supply device data and the model framework to generate a sequence prediction model. As an example, in this embodiment or other embodiments, the self-attention model can be composed of a Transformer.

[0055] S230, predicting the operating status of the power supply equipment in the target area using the sequence prediction model. As an example, in this embodiment or other embodiments, the target area includes but is not limited to the installation area of ​​the power supply equipment and the application area of ​​the power supply equipment.

[0056] It can be seen that this embodiment predicts the operating status of the vehicle power supply equipment by constructing a sequence prediction model, which not only effectively solves the limitations of traditional methods in the face of complex charging scenarios, but also improves the accuracy and stability of the prediction. In addition, this embodiment can realize the real-time deployment of the model and the dynamic update of the prediction results, providing more effective support for the management and scheduling of vehicle power supply equipment.

[0057] According to the above records, in an exemplary embodiment, the process of generating a sequence prediction model in step S220 includes:

[0058] A dual-stream variational autoencoder framework is constructed based on the variational autoencoder, and the charging data is input into the first variational autoencoder in the dual-stream variational autoencoder framework, and the power supply equipment data is input into the second variational autoencoder in the dual-stream variational autoencoder framework. In this embodiment or other embodiments, the dual-stream variational autoencoder framework can introduce probabilistic representation, which can model dynamic and uncertain sequence data; and the dual-stream variational autoencoder framework allows the model to process multivariate information, thereby improving the effectiveness of predicting the occupancy of the vehicle power supply equipment.

[0059] A coding and decoding structure unit is established based on the encoder and decoder in the self-attention model, and the processing result of the first variational self-encoder is output to the first encoder, and the processing result of the second variational self-encoder is output to the second encoder; and the first encoder and the second encoder are used to extract multi-scale information features of the time series, and the multi-scale information features of the time series extracted based on the first encoder are input into the first decoder for time series feature fusion, and the multi-scale information features of the time series extracted based on the second encoder are input into the second decoder for time series feature fusion. In this embodiment or other embodiments, by using the encoder to extract multi-scale information features of the time series, the model has the ability to obtain multi-scale features. And after fusing the feature information of the two time series, the decoder can predict the occupancy of the vehicle power supply equipment.

[0060] An interactive learning structural unit is established, and the processing results of the first variational autoencoder, the processing results of the second variational autoencoder, the time series multi-scale information features extracted by the first encoder, and the time series multi-scale information features extracted by the second encoder are input into the interactive learning structural unit for interactive learning training to obtain interactive learning training results. In this embodiment or other embodiments, by establishing an interactive learning structural unit, a connection between two independent time series feature information can be established, so that the model can perform interactive learning in the latent space. Doing so can improve the model's understanding of the interaction between multiple variables, thereby improving the prediction performance and avoiding the decline in prediction performance caused by ignoring the interaction of multiple variables.

[0061] Inputting the time series multi-scale information features and interactive learning training results from the first encoder and the second encoder into the third decoder simultaneously for time series feature fusion;

[0062] A sequence prediction model is generated based on the time series feature fusion result of the first decoder, the time series feature fusion result of the second decoder, and the time series feature fusion result of the third decoder. In this embodiment or other embodiments, the trained sequence prediction model can also be saved as a model file to facilitate subsequent deployment and use in the prediction stage.

[0063] According to the above records, in an exemplary embodiment, the process of predicting the operating status of the power supply equipment in the target area using the sequence prediction model includes: storing the sequence prediction model as a model file, and deploying it based on the model file to predict the operating status of the power supply equipment in the target area; wherein the deployment environment of the model file includes an application. In this embodiment, the process of predicting the operating status of the power supply equipment in the target area based on the model file deployment includes: converting the model file into a preset file format to obtain a preset file format model; obtaining a folder created in advance or in real time for loading the preset file format model, and installing dependencies in the folder; using a script created in advance or in real time as an inference prediction entry for calling the preset file format model, the script is used to process input data from an external access request, and perform model inference prediction on the local side; creating an application interface service based on the dependency, and receiving an access request fed back by the front-end page, calling the script to load the preset file format model, predicting the operating status of the power supply equipment in the target area, and feeding back the power supply equipment prediction result to the front-end page for display. In this embodiment or other embodiments, the occupancy of the power supply equipment can be displayed by a circular bar chart or a heat map on a map.

[0064] According to the above records, in an exemplary embodiment, the process of acquiring the charging data and power supply equipment data of the target vehicle includes: selecting a collection device according to a data collection type and a data collection target determined in advance or in real time; based on a collection frequency, a collection method and a data format determined in advance or in real time, collecting the charging data and power supply equipment data of the target vehicle through the collection device; performing data cleaning on the charging data and power supply equipment data collected by the collection device, and aggregating the pre-processed charging data and power supply equipment data at the current timestamp according to the time aggregation scale, calculating the occupancy rate of the power supply equipment, and performing random data segmentation; data cleaning includes: removing duplicate data, filling missing values, and correcting erroneous data.

[0065] In another exemplary embodiment of the present invention, Figures 3 to 6 As shown, this embodiment provides a method for predicting electric vehicle power supply equipment, comprising the following steps:

[0066] S1: Data collection and preprocessing. Obtain data in a unified standard interface and use and process it accordingly. Use smart meters to obtain data such as charging power and usage time. Collect the location and power information of the charging pile at regular intervals and save the data in CSV or JSON format. Preprocess the collected data set. The collected raw data may be noisy or incomplete, and data cleaning and processing are required. For example, remove duplicate data, fill in missing values, correct erroneous data, etc.

[0067] S2: Construct a sequence prediction model MTVAE based on variational autoencoder (VAE) and self-attention model Transformer. The sequence prediction model is mainly used to predict the occupancy of electric vehicle power supply equipment under multivariate information marked by time. The entire model structure is mainly divided into three parts: dual-stream variational autoencoder framework, encoding and decoding structure unit and interactive learning structure unit.

[0068] S3: Implement model deployment and prediction based on the sequence prediction model MTVAE, and deploy the model to the application environment so that it can be loaded and used in the application environment.

[0069] S4: Establish an electric vehicle power supply equipment prediction application to predict the occupancy of electric vehicle power supply equipment in the electric vehicle charging pile scenario.

[0070] In this embodiment, the specific process of data collection and preprocessing in step S1 includes:

[0071] S11: Determine data requirements and goals: First, it is necessary to clarify the type of data and goals that need to be collected, such as the location of the charging equipment, power information, frequency of use, etc.

[0072] S12: Select the collection device: Select the appropriate collection device according to the data requirements. Common devices include sensors, GPS positioning devices, smart meters, etc. For example, you can use GPS devices to obtain the location information of charging piles, and use smart meters to obtain data such as charging power and usage time.

[0073] S13: Data collection scheme design: Design a data collection scheme, including collection frequency, collection method (real-time collection or timed collection), data format, etc. For example, the location and power information of the charging pile can be collected at regular intervals and the data can be saved in JSON format.

[0074] S14: Collect data: Start collecting data according to the designed plan. Data collection can be achieved by installing sensors or connecting smart devices. Ensure that the collection equipment is operating normally and record any abnormalities or problems during the collection process.

[0075] S15: Data cleaning and processing: The collected raw data may be noisy or incomplete, so data cleaning and processing are required. Duplicate data should be removed, missing values ​​should be filled, erroneous data should be corrected, and the EV power supply equipment information of the previous three days at the current timestamp should be aggregated according to the time aggregation scale, the corresponding EV power supply equipment occupancy rate should be calculated, and the data should be randomly split for model use.

[0076] S16: Data storage and management: The cleaned data is stored in a database or file, and a data management system is established. Access permissions and backup strategies can be set as needed to ensure data security and reliability. The database is managed by MySQL, and a reasonable database architecture is designed to store and manage the data of electric vehicle charging stations. The database not only contains the basic information of the charging station, but also designs a complex relational model to store information such as the usage records of the charging station, user evaluations, and detailed specifications of the charging equipment. This design makes data storage more systematic and standardized, which is convenient for subsequent data analysis and model training.

[0077] S17: Data analysis and application: Use the collected data for analysis and application to verify the effectiveness and advancement of the proposed method.

[0078] In this embodiment, the specific process of constructing the sequence prediction model MTVAE based on variational autoencoder VAE and Transformer in step S2 includes:

[0079] S21: Constructing a dual-stream variational autoencoder framework: This framework introduces probabilistic representation and is able to model dynamic, uncertain sequence data. It allows the model to process multivariate information and improves the effectiveness of predicting the occupancy of electric vehicle power supply equipment.

[0080] S22: Establishing a coding and decoding structure unit: This unit consists of a first encoder E1, a second encoder E2, a first decoder D1, a second decoder D2 and a third decoder D3. The encoder is responsible for extracting multi-scale information features of the time series, so that the model has the ability to obtain multi-scale features. The decoder predicts the occupancy of the electric vehicle power supply equipment after fusing the feature information of the two time series.

[0081] S23: Establish interactive learning structural unit: This part is responsible for establishing the connection between the feature information of two independent time series, so that the model can interactively learn in the latent space. Doing so can improve the model's understanding of the interaction between multiple variables, thereby improving the prediction performance and avoiding the decline in prediction performance caused by ignoring the interaction of multiple variables.

[0082] S24: Model training: Use the acquired data set to train the constructed MTVAE model. During the training process, the model will learn how to capture dynamic and uncertain sequence data and improve the prediction accuracy of the occupancy of electric vehicle power supply equipment.

[0083] S25: Model saving: After the model training is completed, the trained MTVAE model is saved as a model file for use in the deployment and prediction stages.

[0084] In this embodiment, in step S3, model deployment and prediction are implemented based on the sequence prediction model MTVAE. The specific process of deploying the model to the application environment so as to load and use it in the application environment includes:

[0085] S31: Convert the model file to ONNX (Open Neural Network Exchange) format: Use the torch tool to convert the saved model file to ONNX format for inference on the front-end page.

[0086] S32: Create a model loading folder: Create a folder locally for loading the model and install necessary dependencies, including Numpy, Flask (for creating API services), and ONNX Python library.

[0087] S33: Write Python script: Write a Python script as the entry point for calling the ONNX model for inference. The script is responsible for preprocessing the input data from the HTTP request and performing model inference using local computing resources.

[0088] S34: Model reasoning: In the Python script, the loaded ONNX model is used for reasoning to obtain the probability that the electric vehicle power supply device is occupied. According to the set threshold, it is determined whether the device is considered to be occupied, and the result is converted into JSON format. In this embodiment, the threshold can be set according to the actual situation, for example, it can be set to 0.7.

[0089] S35: API service: Use Flask or a similar framework to create an API service to receive HTTP requests from the front-end page and call the above Python script for model inference and result return.

[0090] S36: Update page content in real time: Based on the model prediction results, front-end developers can easily update page content and provide real-time feedback to users, such as displaying the probability of device occupancy or visualization of prediction results.

[0091] In this embodiment, in step S4, an electric vehicle power supply equipment prediction application is established, and the process of predicting the occupancy of the electric vehicle power supply equipment in the electric vehicle charging pile scenario includes:

[0092] S41: Data loading module design: Design a data loading module to load the status information and other relevant data of the electric vehicle charging pile from the collected data. This module is combined with the data collection and preprocessing step S1 to ensure that the data required for display and prediction can be effectively obtained.

[0093] S42: Model prediction module design: Design a model prediction module to call the model deployed in step S3 to predict the occupancy of electric vehicle power supply equipment. The module can receive requests from the front-end page, call the model for reasoning, and return the prediction results.

[0094] S43: Front-end rendering module design: Design a front-end rendering module to display the model prediction results to the user in a visual way. This module can update the page content in real time and display the occupancy status of the electric vehicle power supply equipment, such as through a circular bar chart or a heat map on the map to show the occupancy status of the equipment.

[0095] S44: Based on the designs of the data loading module, the model prediction module and the front-end rendering module, an electric vehicle power supply equipment prediction application based on a sequence model is integrated.

[0096] It can be seen that this embodiment uses advanced sequence prediction technology, combined with historical data and real-time information, to build a more accurate and reliable prediction model. Specifically, this embodiment uses a sequence prediction model MTVAE based on a variational autoencoder and a self-attention model Transformer, and by constructing key components such as a dual-stream variational autoencoder framework, a codec structure unit, and an interactive learning structure unit, it achieves efficient prediction of the occupancy of electric vehicle power supply equipment. This embodiment can not only process multivariate information and dynamic changes, but also achieve better prediction results in practical applications. At the same time, this embodiment also takes into account the actual needs of model deployment and application. By converting the model into ONNX format and using tools such as Flask to implement model deployment and API services, it achieves efficient deployment of the model and real-time update of prediction results, providing strong support for the management and scheduling of power supply equipment.

[0097] In summary, the present invention provides a vehicle power supply equipment prediction method, firstly, the charging data and power supply equipment data of the target vehicle are obtained; wherein, the charging data includes the charging power and the charging time, the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle; then, a model framework is constructed based on a variational autoencoder and a self-attention model, and the charging data, the power supply equipment data and the model framework are used for training to generate a sequence prediction model; finally, the sequence prediction model is used to predict the operating status of the power supply equipment in the target area. It can be seen that the method predicts the operating status of the vehicle power supply equipment by constructing a sequence prediction model, which not only effectively solves the limitations of the traditional method in the face of complex charging scenarios, but also improves the accuracy and stability of the prediction. The method can also realize the real-time deployment of the model and the dynamic update of the prediction results, providing more effective support for the management and scheduling of the vehicle power supply equipment. In addition, the method can also use a sequence prediction model MTVAE based on a variational autoencoder and a self-attention model Transformer to construct a sequence prediction model, and by constructing key components such as a dual-stream variational autoencoder framework, a coding and decoding structure unit and an interactive learning structure unit, an efficient prediction of the occupancy of the electric vehicle power supply equipment is achieved. This method can not only handle multivariate information and dynamic changes, but also achieve better prediction results in practical applications. At the same time, this method also takes into account the actual needs of model deployment and application. By converting the model into ONNX format and using tools such as Flask to implement model deployment and API services, it achieves efficient deployment of the model and real-time update of prediction results, providing strong support for the management and scheduling of power supply equipment.

[0098] In another exemplary embodiment of the present invention, Figure 7 As shown, this embodiment also provides a vehicle power supply equipment prediction system, including:

[0099] The data acquisition module 710 is used to obtain the charging data and power supply equipment data of the target vehicle; wherein the charging data includes the charging power and the charging time, and the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle. As an example, in this embodiment or other embodiments, the target vehicle includes but is not limited to a new energy vehicle, and the power supply equipment includes but is not limited to a charging pile. For example, in this embodiment, the new energy vehicle may be an electric vehicle.

[0100] The model training module 720 is used to build a model framework based on the variational autoencoder and the self-attention model, and train using the charging data, the power supply device data and the model framework to generate a sequence prediction model. As an example, in this embodiment or other embodiments, the self-attention model can be composed of a Transformer.

[0101] The prediction module 730 is used to predict the operating status of the power supply equipment in the target area by using the sequence prediction model. As an example, in this embodiment or other embodiments, the target area includes but is not limited to the installation area of ​​the power supply equipment and the application area of ​​the power supply equipment.

[0102] It can be seen that this embodiment predicts the operating status of the vehicle power supply equipment by constructing a sequence prediction model, which not only effectively solves the limitations of traditional methods in the face of complex charging scenarios, but also improves the accuracy and stability of the prediction. In addition, this embodiment can realize the real-time deployment of the model and the dynamic update of the prediction results, providing more effective support for the management and scheduling of vehicle power supply equipment.

[0103] According to the above records, in an exemplary embodiment, the process of the model training module 720 generating a sequence prediction model includes:

[0104] A dual-stream variational autoencoder framework is constructed based on the variational autoencoder, and the charging data is input into the first variational autoencoder in the dual-stream variational autoencoder framework, and the power supply equipment data is input into the second variational autoencoder in the dual-stream variational autoencoder framework. In this embodiment or other embodiments, the dual-stream variational autoencoder framework can introduce probabilistic representation, which can model dynamic and uncertain sequence data; and the dual-stream variational autoencoder framework allows the model to process multivariate information, thereby improving the effectiveness of predicting the occupancy of the vehicle power supply equipment.

[0105] A coding and decoding structure unit is established based on the encoder and decoder in the self-attention model, and the processing result of the first variational self-encoder is output to the first encoder, and the processing result of the second variational self-encoder is output to the second encoder; and the first encoder and the second encoder are used to extract multi-scale information features of the time series, and the multi-scale information features of the time series extracted based on the first encoder are input into the first decoder for time series feature fusion, and the multi-scale information features of the time series extracted based on the second encoder are input into the second decoder for time series feature fusion. In this embodiment or other embodiments, by using the encoder to extract multi-scale information features of the time series, the model has the ability to obtain multi-scale features. And after fusing the feature information of the two time series, the decoder can predict the occupancy of the vehicle power supply equipment.

[0106] An interactive learning structural unit is established, and the processing results of the first variational autoencoder, the processing results of the second variational autoencoder, the time series multi-scale information features extracted by the first encoder, and the time series multi-scale information features extracted by the second encoder are input into the interactive learning structural unit for interactive learning training to obtain interactive learning training results. In this embodiment or other embodiments, by establishing an interactive learning structural unit, a connection between two independent time series feature information can be established, so that the model can perform interactive learning in the latent space. Doing so can improve the model's understanding of the interaction between multiple variables, thereby improving the prediction performance and avoiding the decline in prediction performance caused by ignoring the interaction of multiple variables.

[0107] Inputting the time series multi-scale information features and interactive learning training results from the first encoder and the second encoder into the third decoder simultaneously for time series feature fusion;

[0108] A sequence prediction model is generated based on the time series feature fusion result of the first decoder, the time series feature fusion result of the second decoder, and the time series feature fusion result of the third decoder. In this embodiment or other embodiments, the trained sequence prediction model can also be saved as a model file to facilitate subsequent deployment and use in the prediction stage.

[0109] According to the above records, in an exemplary embodiment, the process of predicting the operating status of the power supply equipment in the target area using the sequence prediction model includes: storing the sequence prediction model as a model file, and deploying it based on the model file to predict the operating status of the power supply equipment in the target area; wherein the deployment environment of the model file includes an application. In this embodiment, the process of predicting the operating status of the power supply equipment in the target area based on the model file deployment includes: converting the model file into a preset file format to obtain a preset file format model; obtaining a folder created in advance or in real time for loading the preset file format model, and installing dependencies in the folder; using a script created in advance or in real time as an inference prediction entry for calling the preset file format model, the script is used to process input data from an external access request, and perform model inference prediction on the local side; creating an application interface service based on the dependency, and receiving an access request fed back by the front-end page, calling the script to load the preset file format model, predicting the operating status of the power supply equipment in the target area, and feeding back the power supply equipment prediction result to the front-end page for display. In this embodiment or other embodiments, the occupancy of the power supply equipment can be displayed by a circular bar chart or a heat map on a map.

[0110] According to the above records, in an exemplary embodiment, the process of acquiring the charging data and power supply equipment data of the target vehicle includes: selecting a collection device according to a data collection type and a data collection target determined in advance or in real time; based on a collection frequency, a collection method and a data format determined in advance or in real time, collecting the charging data and power supply equipment data of the target vehicle through the collection device; performing data cleaning on the charging data and power supply equipment data collected by the collection device, and aggregating the pre-processed charging data and power supply equipment data at the current timestamp according to the time aggregation scale, calculating the occupancy rate of the power supply equipment, and performing random data segmentation; data cleaning includes: removing duplicate data, filling missing values, and correcting erroneous data.

[0111] It should be noted that the vehicle power supply equipment prediction system provided in the above embodiment and the vehicle power supply equipment prediction method provided in the above embodiment belong to the same concept, wherein the specific manner in which the vehicle power supply equipment prediction method performs operations has been described in detail in the above method embodiment, and will not be repeated here. In actual applications, the vehicle power supply equipment prediction system provided in the above embodiment can allocate the above functions to different functional modules as needed, that is, divide the internal structure of the vehicle power supply equipment prediction system into different functional modules, and then implement all or part of the functions of the corresponding functional modules through the vehicle power supply equipment prediction method described in the above embodiment, and this is not specifically limited here.

[0112] In summary, the present invention provides a vehicle power supply equipment prediction system, which first obtains the charging data and power supply equipment data of the target vehicle; wherein the charging data includes the charging power and the charging time, and the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle; then, a model framework is constructed based on a variational autoencoder and a self-attention model, and the charging data, the power supply equipment data and the model framework are used for training to generate a sequence prediction model; finally, the sequence prediction model is used to predict the operating status of the power supply equipment in the target area. It can be seen that the system predicts the operating status of the vehicle power supply equipment by constructing a sequence prediction model, which not only effectively solves the limitations of traditional methods in the face of complex charging scenarios, but also improves the accuracy and stability of the prediction. The system can also realize the real-time deployment of the model and the dynamic update of the prediction results, providing more effective support for the management and scheduling of the vehicle power supply equipment. In addition, the system can also use a sequence prediction model MTVAE based on a variational autoencoder and a self-attention model Transformer, and by constructing key components such as a dual-stream variational autoencoder framework, a codec structure unit and an interactive learning structure unit, an efficient prediction of the occupancy of the electric vehicle power supply equipment is achieved. This system can not only process multivariate information and dynamic changes, but also achieve better prediction results in practical applications. At the same time, this system also takes into account the actual needs of model deployment and application. By converting the model into ONNX format and using tools such as Flask to implement model deployment and API services, it achieves efficient model deployment and real-time update of prediction results, providing strong support for the management and scheduling of power supply equipment.

[0113] In another embodiment of the present invention, the embodiment further provides a vehicle networking control device, which may include a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program so that the vehicle networking control device executes Figure 2 and / or Figure 3 The steps of the vehicle power supply equipment prediction method. Figure 8 1 shows a schematic diagram of the structure of a vehicle networking control device 1000. Figure 8 As shown, the Internet of Vehicles control device 1000 includes: a processor 1010 , a memory 1020 , a power supply 1030 , a display unit 1040 , and an input unit 1060 .

[0114] The processor 1010 is the control center of the vehicle networking control device 1000. It uses various interfaces and lines to connect various components, and executes various functions of the vehicle networking control device 1000 by running or executing computer programs / instructions stored in the memory 1020, thereby monitoring the vehicle networking control device 1000 as a whole. In this embodiment of the present invention, when the processor 1010 calls the computer program stored in the memory 1020, the following is executed: Figure 2 and / or Figure 3 The steps of the vehicle power supply equipment prediction method. Optionally, the processor 1010 may include one or more processing units; preferably, the processor 1010 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface and application, etc., and the modem processor mainly processes wireless communication. In some embodiments, the processor, memory, and memory may be implemented on a single chip, and in some embodiments, they may also be implemented separately on separate chips.

[0115] The memory 1020 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, various applications, etc.; the data storage area may store instruction data created according to the use of the vehicle networking control device 1000, etc. In addition, the memory 1020 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage devices, etc.

[0116] The Internet of Vehicles control device 1000 also includes a power source 1030 (such as a battery) for supplying power to various components. The power source can be logically connected to the processor 1010 through a power management system, thereby managing functions such as charging, discharging, and power consumption through the power management system.

[0117] The display unit 1040 may be used to display information input by the user or information provided to the user and various menus of the vehicle networking control device 1000, and in the embodiment of the present invention, it is mainly used to display the display interface of each application in the vehicle networking control device 1000 and objects such as text and pictures displayed in the display interface. The display unit 1040 may include a display panel 1050. The display panel 1050 may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), etc.

[0118] The input unit 1060 may be used to receive information such as numbers or characters input by the user. The input unit 1060 may include a touch panel 1070 and other input devices 1080. The touch panel 1070, also known as a touch screen, may collect touch operations of the user on or near it (such as operations of the user using a finger, a stylus, or any other suitable object or accessory on or near the touch panel 1070).

[0119] Specifically, the touch panel 1070 can detect the user's touch operation, detect the signal brought by the touch operation, convert these signals into touch point coordinates, send them to the processor 1010, and receive and execute the command sent by the processor 1010. In addition, the touch panel 1070 can be implemented in various types such as resistive, capacitive, infrared and surface acoustic wave. Other input devices 1080 can include but are not limited to one or more of a physical keyboard, function keys (such as volume control keys, power on / off keys, etc.), a trackball, a mouse, a joystick, etc.

[0120] Of course, the touch panel 1070 can cover the display panel 1050. When the touch panel 1070 detects a touch operation on or near it, it is transmitted to the processor 1010 to determine the type of touch event, and then the processor 1010 provides a corresponding visual output on the display panel 1050 according to the type of touch event. Figure 8 In the embodiment, the touch panel 1070 and the display panel 1050 are two independent components to realize the input and output functions of the Internet of Vehicles control device 1000 , but in some embodiments, the touch panel 1070 and the display panel 1050 can be integrated to realize the input and output functions of the Internet of Vehicles control device 1000 .

[0121] The Internet of Vehicles control device 1000 may further include one or more sensors, such as a pressure sensor, a gravity acceleration sensor, a proximity light sensor, etc. Of course, according to the needs of specific applications, the Internet of Vehicles control device 1000 may also include other components such as a camera.

[0122] The embodiment of the present invention further provides a computer-readable storage medium, in which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the above-mentioned device can perform the present invention as follows: Figure 2 and / or Figure 3 The steps of the vehicle power supply equipment prediction method.

[0123] It can be understood by those skilled in the art that Figure 8This is only an example of a vehicle networking control device and does not constitute a limitation on the device. The device may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For the convenience of description, the above parts are divided into modules (or units) according to their functions and described separately. Of course, when implementing the present invention, the functions of each module (or unit) can be implemented in the same or multiple software or hardware.

[0124] Those skilled in the art should understand that the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be applied to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the functions specified in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0125] In another exemplary embodiment of the present invention, the embodiment further provides a vehicle, which includes a vehicle power supply equipment prediction system as described in the above embodiment, or includes a vehicle network control device as described in the above embodiment. It should be noted that since the specific manner in which the vehicle power supply equipment prediction system and the vehicle network control device perform operations has been described in detail in the embodiments, the technical functions and effects of the vehicle provided in this embodiment can be referred to the above embodiments, and will not be repeated here.

[0126] It should be noted that the above embodiments collect, store, use, process, transmit, provide, disclose, delete, etc. related data (such as charging data and power supply equipment data, etc.) with or with the consent of the user. For example, the charging data and power supply equipment data are authorized with the knowledge and consent of the user; or the user actively provides it after reading the relevant instructions, or the user actively authorizes / provides / uploads it when using some or all of the functions described in the above embodiments, or other methods / channels are obtained with or with the consent of the user.

[0127] It should be understood that, although the terms first, second, third, etc. may be used to describe encoders, etc. in embodiments of the present invention, these terms are only used to distinguish encoders from each other. For example, without departing from the scope of embodiments of the present invention, a first encoder may also be referred to as a second encoder, and similarly, a second encoder may also be referred to as a first encoder.

[0128] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A vehicle power supply equipment prediction method, characterized in that: The method comprises the following steps: Obtain charging data and power supply equipment data of the target vehicle; wherein the charging data includes charging power and charging time, and the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle; Building a model framework based on a variational autoencoder and a self-attention model, and using the charging data, the power supply device data, and the model framework for training to generate a sequence prediction model; The sequence prediction model is used to predict the operating status of the power supply equipment in the target area.

2. The vehicle power supply equipment prediction method according to claim 1, characterized in that: The process of building a sequence prediction model includes: A dual-stream variational autoencoder framework is constructed based on the variational autoencoder, and the charging data is input into a first variational autoencoder in the dual-stream variational autoencoder framework, and the power supply device data is input into a second variational autoencoder in the dual-stream variational autoencoder framework; A coding and decoding structure unit is established based on the encoder and decoder in the self-attention model, and the processing result of the first variational autoencoder is output to the first encoder, and the processing result of the second variational autoencoder is output to the second encoder; and the first encoder and the second encoder are used to extract time series multi-scale information features, and the time series multi-scale information features extracted based on the first encoder are input into the first decoder for time series feature fusion, and the time series multi-scale information features extracted based on the second encoder are input into the second decoder for time series feature fusion; Establishing an interactive learning structure unit, and inputting the processing result of the first variational autoencoder, the processing result of the second variational autoencoder, the time series multi-scale information features extracted by the first encoder, and the time series multi-scale information features extracted by the second encoder into the interactive learning structure unit for interactive learning training to obtain an interactive learning training result; Inputting the time series multi-scale information features from the first encoder and the second encoder and the interactive learning training results into the third decoder simultaneously for time series feature fusion; The sequence prediction model is generated according to the time series feature fusion result of the first decoder, the time series feature fusion result of the second decoder, and the time series feature fusion result of the third decoder.

3. The vehicle power supply equipment prediction method according to claim 1 or 2, characterized in that: The process of using the sequence prediction model to predict the operating status of the power supply equipment in the target area includes: storing the sequence prediction model as a model file, and deploying it based on the model file to predict the operating status of the power supply equipment in the target area; wherein the deployment environment of the model file includes an application.

4. The vehicle power supply equipment prediction method according to claim 3, characterized in that: The process of deploying based on the model file and predicting the operating status of the power supply equipment in the target area includes: Converting the model file into a preset file format to obtain a preset file format model; Obtain a folder created in advance or in real time for loading the preset file format model, and install dependencies in the folder; Using a script created in advance or in real time as an inference prediction entry for calling the preset file format model, the script is used to process input data from an external access request and perform model inference prediction on a local end; An application program interface service is created based on the dependencies, and an access request fed back by a front-end page is received, the script is called to load the preset file format model, the operating status of the power supply equipment in the target area is predicted, and the power supply equipment prediction result is fed back to the front-end page for display.

5. The vehicle power supply equipment prediction method according to claim 1 or 2, characterized in that: The process of obtaining the charging data of the target vehicle and the data of the power supply equipment includes: Select the collection equipment according to the data collection type and data collection target determined in advance or in real time; Based on a collection frequency, collection method and data format determined in advance or in real time, the charging data of the target vehicle and the power supply device data are collected by the collection device; The charging data and power supply equipment data collected by the collection device are cleaned, and the pre-processed charging data and power supply equipment data at the current timestamp are aggregated according to the time aggregation scale, and the occupancy rate of the power supply equipment is calculated, and the data is randomly segmented; the data cleaning includes: removing duplicate data, filling missing values, and correcting erroneous data.

6. The vehicle power supply equipment prediction method according to claim 1 or 2, characterized in that: The target vehicle includes a new energy vehicle, and the power supply equipment includes a charging pile.

7. A vehicle power supply equipment prediction system, characterized in that: The system comprises: A data acquisition module, used to obtain charging data and power supply equipment data of the target vehicle; wherein the charging data includes charging power and charging time, and the power supply equipment data includes the location of the power supply equipment and the power of the power supply equipment, and the power supply equipment is used to supply power to the target vehicle; A model training module, used to build a model framework based on a variational autoencoder and a self-attention model, and to train using the charging data, the power supply device data, and the model framework to generate a sequence prediction model; The prediction module is used to predict the operating status of the power supply equipment in the target area by using the sequence prediction model.

8. The vehicle power supply equipment prediction system according to claim 7, characterized in that: The process of generating a sequence prediction model by the model training module includes: A dual-stream variational autoencoder framework is constructed based on the variational autoencoder, and the charging data is input into a first variational autoencoder in the dual-stream variational autoencoder framework, and the power supply device data is input into a second variational autoencoder in the dual-stream variational autoencoder framework; A coding and decoding structure unit is established based on the encoder and decoder in the self-attention model, and the processing result of the first variational autoencoder is output to the first encoder, and the processing result of the second variational autoencoder is output to the second encoder; and the first encoder and the second encoder are used to extract time series multi-scale information features, and the time series multi-scale information features extracted based on the first encoder are input into the first decoder for time series feature fusion, and the time series multi-scale information features extracted based on the second encoder are input into the second decoder for time series feature fusion; Establishing an interactive learning structure unit, and inputting the processing result of the first variational autoencoder, the processing result of the second variational autoencoder, the time series multi-scale information features extracted by the first encoder, and the time series multi-scale information features extracted by the second encoder into the interactive learning structure unit for interactive learning training to obtain an interactive learning training result; Inputting the time series multi-scale information features from the first encoder and the second encoder and the interactive learning training results into the third decoder simultaneously for time series feature fusion; The sequence prediction model is generated according to the time series feature fusion result of the first decoder, the time series feature fusion result of the second decoder, and the time series feature fusion result of the third decoder.

9. A vehicle networking control device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the vehicle power supply equipment prediction method according to any one of claims 1 to 6.

10. A vehicle, characterized in that: The vehicle includes the vehicle power supply equipment prediction system as described in claim 7 or 8, or includes the vehicle networking control device as described in claim 9.

Images