Multi-source data fusion electric vehicle endurance dynamic optimization system and management method

Abstract

The invention discloses an electric vehicle endurance dynamic optimization system and management method based on multi-source data fusion, and the system comprises a vehicle data acquisition module, an energy consumption prediction module, a dynamic optimization module and a feedback module: the vehicle data acquisition module is used for collecting and preprocessing multi-source data of an electric vehicle; the energy consumption prediction module is used for predicting the endurance of the electric vehicle according to the multi-source data acquired by the vehicle data acquisition module and predicting the residual endurance distance; the dynamic optimization module is used for evaluating the charging stations on the planned route of the electric vehicle according to the prediction result of the energy consumption prediction module and generating a charging station recommendation sequence according to the evaluation result; the feedback module is used for feeding back to the user according to the charging station recommendation sequence of the dynamic optimization module; according to the method, the accuracy of endurance calculation is improved by combining the actual terrain; and furthermore, by introducing the electricity price and the queuing time cost, the practicability of the charging recommendation scheme is improved, and the user experience is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy management for new energy vehicles, and in particular to a multi-source data fusion dynamic optimization system and management method for electric vehicle endurance. Background Art

[0002] As market conditions and user demands continue to evolve, electric vehicles, with their significant advantages such as zero tailpipe emissions and low noise levels, have become a core driver of the automotive industry's transition to a green, low-carbon future. In recent years, global electric vehicle sales have continued to climb, with countries implementing policies to support the industry's development and consumers increasingly focused on environmentally friendly mobility. This has ushered in unprecedented growth opportunities for the electric vehicle market. However, alongside this rapid growth, battery life-related issues within the electric vehicle user experience have become increasingly prominent, hindering further industry breakthroughs.

[0003] Traditional range prediction methods rely primarily on historical energy consumption data, analyzing and modeling energy consumption information accumulated during past driving to estimate a vehicle's remaining range. However, this approach has significant flaws. In real-world driving, terrain factors significantly impact electric vehicle energy consumption. On hilly roads, the vehicle must overcome gravity to generate work, requiring the motor to output greater power, significantly increasing energy consumption. While energy recovery technology can be used to replenish power on long downhill sections, the efficiency of energy recovery varies depending on the slope, length, and driving conditions. Traditional prediction models based on historical energy consumption data are unable to perceive the current topography of the route in real time, resulting in significant deviations between range predictions and actual conditions, complicating user journey planning. When it comes to charging station recommendations, most existing systems only consider charging station location information, recommending nearby stations. This single-dimensional approach fails to factor in electricity price fluctuations and queue time costs, making it difficult to provide users with the most economical and efficient charging solutions.

[0004] Therefore, there is an urgent need for an electric vehicle endurance dynamic optimization system and management method that integrates multi-source data to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a multi-source data fusion electric vehicle endurance dynamic optimization system and management method to solve the problems raised in the above background technology.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] The electric vehicle endurance dynamic optimization system based on multi-source data fusion includes a vehicle data acquisition module, an energy consumption prediction module, a dynamic optimization module, and a feedback module:

[0008] The vehicle data acquisition module is used to collect multi-source data of the electric vehicle, including vehicle parameter data, wind speed data and driving behavior data, and perform pre-processing;

[0009] The energy consumption prediction module is used to predict the cruising range of the electric vehicle and the remaining cruising distance based on the multi-source data acquired by the vehicle data acquisition module;

[0010] The dynamic optimization module includes an electricity price prediction unit, a queue prediction unit, and an evaluation unit, which is used to evaluate the charging stations on the planned route of the electric vehicle based on the prediction results of the energy consumption prediction module and generate a recommended sequence of charging stations based on the evaluation results;

[0011] The feedback module is used to provide feedback to the user based on the charging station recommendation sequence of the dynamic optimization module.

[0012] Furthermore, the vehicle data acquisition module includes a data acquisition unit and a data preprocessing unit:

[0013] The data acquisition unit connects to the map software through the API based on the real-time GPS coordinates of the vehicle and parses the road topology data returned by the map service. In combination with the on-board sensors, the on-board network and the external environment monitoring equipment, the data acquisition unit obtains the vehicle parameter data of the electric vehicle in real time. The vehicle parameter data includes battery parameters and operating parameters. The battery parameters include battery state of charge, battery temperature, battery internal resistance and battery health status; the operating parameters include vehicle speed, acceleration, mileage and road gradient. The data acquisition unit connects to the national meteorological data platform through the HTTPS protocol, generates a geographic grid code based on the vehicle location, requests real-time meteorological observation data within a preset range, extracts the wind speed vector and obtains the wind speed data. The data acquisition unit directly reads the original signal of the on-board diagnostic system through the CAN bus monitor to obtain driving behavior data in real time.

[0014] The data preprocessing unit is used to synchronize the vehicle parameter data, wind speed data and driving behavior data collected by the data acquisition unit; perform timestamp alignment operations through the GPS clock and PTP protocol, and periodically calibrate the vehicle parameter data, wind speed data and driving behavior data.

[0015] It should be explained that multi-dimensional data of the vehicle's operation process is obtained in real time through on-board sensors, on-board networks, and external environmental monitoring equipment. Among them, vehicle battery-related data includes battery state of charge (SOC), battery temperature, battery internal resistance, battery state of health (SOH), etc. These parameters directly reflect the battery's performance and remaining power; vehicle operating parameters include speed, acceleration, mileage, road gradient, etc., which can reflect the vehicle's driving status and working conditions; environmental data includes ambient temperature and humidity, wind speed and direction, etc. Environmental factors have an impact on vehicle energy consumption that cannot be ignored; in addition, the system also collects driving behavior data, such as rapid acceleration frequency, rapid deceleration frequency, average speed changes, etc. These data can reflect the driver's driving habits, thereby affecting vehicle energy consumption.

[0016] Because the collected data may contain noise, outliers, or inconsistent formats, the preprocessing unit cleans and standardizes the raw data. Using algorithms such as data filtering and outlier detection and correction, it removes noise and erroneous information from the data. It also converts data of different types and formats into a unified format for analysis and processing by subsequent modules.

[0017] Furthermore, the energy consumption prediction module includes:

[0018] A dynamic energy consumption model is constructed using a long short-term memory (LSTM) neural network. The input layer receives preprocessed vehicle parameter data, wind speed data, and driving behavior data in a time series format. Each time step corresponds to a set of data containing battery parameters, operating parameters, wind speed, and driving behavior parameters. Multiple LSTM layers are set up, and a fully connected layer integrates the feature vectors output by the LSTM layers. The output layer calculates and outputs the vehicle's remaining range based on the results of the fully connected layer.

[0019] Extract the working log of the electric vehicle, extract historical vehicle parameter data, wind speed data, driving behavior data and the remaining range of the vehicle, and divide the historical data into training set and validation set according to the specified ratio;

[0020] The dynamic energy consumption model is trained using the training set, and the mean square error is used as the loss function to measure the error between the remaining cruising distance predicted by the model and the actual remaining cruising distance. Based on the validation set, the model parameters are iteratively updated using stochastic gradient descent and its improved algorithm, and the weights and biases in the model are continuously adjusted until the loss function value is less than or equal to a preset loss threshold. The iterative operation is stopped to obtain a trained dynamic energy consumption model.

[0021] The preprocessed real-time multi-source data is input into the trained dynamic energy consumption model in sequence according to time steps, and the predicted value of the vehicle's remaining cruising distance in the current state is output in the output layer.

[0022] It's important to explain that the energy consumption prediction module is one of the system's core modules. Its primary function is to accurately predict the vehicle's remaining range based on preprocessed data provided by the vehicle data acquisition module. This module uses a long short-term memory (LSTM) neural network to construct a dynamic energy consumption model. LSTM neural networks have powerful time-series data processing capabilities, effectively capturing the dynamic impact of various factors on energy consumption during vehicle operation. The model inputs preprocessed data such as battery SOC, temperature, vehicle speed, acceleration, ambient temperature and humidity, and road gradient. After learning and calculating, the LSTM neural network outputs the vehicle's remaining range. To ensure accurate and timely predictions, the system updates the predictions every five minutes. Furthermore, the energy consumption coefficient is dynamically adjusted based on real-time traffic information, such as traffic congestion and changes in road gradient. For example, on congested roads, where vehicles frequently start and stop, energy consumption can increase significantly. The system promptly adjusts the energy consumption coefficient based on real-time traffic data to ensure that the remaining range prediction is more accurate.

[0023] Furthermore, the energy consumption prediction module also includes updating the remaining endurance prediction value once every time node.

[0024] Furthermore, the electricity price prediction unit includes:

[0025] Based on the planned path of the electric vehicle and the remaining cruising distance, obtaining a maximum cruising path of the electric vehicle;

[0026] Based on the maximum endurance path, all charging stations on the maximum endurance path are obtained to construct a charging station set;

[0027] For charging station A in the charging station set i To make electricity price forecasts:

[0028] By analyzing the historical load data, power generation cost data, and power market transaction data of the power grid along the planned path and combining it with a time series prediction algorithm, the power price trend of the power grid in the future period is predicted.

[0029] Based on charging station A i The control coefficient of the charging station is calculated based on the historical electricity price data and the grid electricity price data of the corresponding period;

[0030] Based on the electricity price trend of the power grid in the future period, the electricity price trend of the charging station in the future period is obtained.

[0031] It's important to explain that by analyzing historical grid load data, generation cost data, and electricity market transaction data, combined with a time series prediction algorithm, we predict electricity price trends for the next two hours at each charging station along the planned route. For example, during peak hours, when the grid load is high, electricity prices may rise; while during low-demand nighttime hours, electricity prices are relatively low. By accurately predicting electricity price trends, users can optimize their charging schedules and reduce costs.

[0032] Furthermore, the queue prediction unit includes:

[0033] Connect to charging station A i Real-time occupancy data, including the usage status of charging piles and the number of vehicles currently queuing, and statistics on the number of charging piles in working condition;

[0034] Based on the queuing theory model, input charging station A i Output the number of normal charging piles, the number of charging piles in working state, the number of vehicles in queue and the historical average service time of charging station A i The expected waiting time.

[0035] It is important to explain that the real-time occupancy data of connected charging stations, including the usage status of charging piles and the number of vehicles currently in the queue, is used to calculate the expected waiting time at each charging station based on queuing theory models, such as the M / M / c queuing model. The M / M / c queuing model is a classic queuing theory model, where "M" represents that both the arrival and service processes follow an exponential distribution, and "c" represents the number of charging piles. By inputting parameters such as the number of charging piles, the current number of vehicles in the queue, and the historical average service time, the model can accurately predict user waiting times at each charging station, helping users avoid queue peaks and improve charging efficiency.

[0036] Furthermore, the assessment units include:

[0037] According to charging station A i The location data and the average speed of the electric vehicle are used to calculate the speed at which the vehicle reaches charging station A. i duration;

[0038] Based on charging station A i The estimated waiting time of charging station A is calculated i The actual waiting time;

[0039] Construct a multi-objective evaluation function Score = α*(the electricity price at the time of the expected waiting time) + β*(the actual waiting time) + γ*(the charging station A i The distance from the maximum endurance path) for charging station A iPerform a comprehensive evaluation; Score represents the comprehensive score, α, β, and γ represent weights, and α, β, and γ support user-defined configuration;

[0040] Perform a comprehensive evaluation on all charging stations in the charging station set, calculate the comprehensive score of each charging station, and sort the charging stations from high to low according to the comprehensive score to generate a recommended charging station sequence.

[0041] It should be explained that users can adjust the weights of various indicators according to their own needs, such as paying more attention to economic costs (increasing the α value) or paying more attention to time efficiency (increasing the β value). The system will calculate the comprehensive score of each charging station based on the weight set by the user, and sort the charging stations according to the scores; by inputting parameters such as the number of charging piles, the current number of vehicles in the queue, and the historical average service time, the model can accurately predict the waiting time of users at each charging station, help users avoid queuing peaks, and improve charging efficiency.

[0042] Furthermore, the feedback module includes:

[0043] When the vehicle's remaining range falls below the preset range threshold, the feedback module will automatically trigger a voice warning function to remind the user to charge in time; it will mark recommended charging stations on the navigation map, display the estimated remaining power upon arrival at each station, and a comprehensive scoring matrix for each station.

[0044] It should be explained that the feedback module is responsible for presenting the system's analysis results to users in an intuitive and convenient manner, supporting their charging decisions. Its main functions include:

[0045] Generation and display of charging station priority lists: Based on the comprehensive scores of charging stations calculated by the dynamic optimization module, a charging station priority list is generated, and the three best charging stations and their evaluation parameters, such as electricity price, estimated waiting time, distance, etc., are displayed to users through the vehicle head-up display (HUD) system. Users can quickly obtain key information without having to look down at their mobile phones or vehicle screens, thereby improving driving safety. Battery life warning and navigation: When the vehicle's remaining battery life is lower than the preset threshold, the feedback module will automatically trigger the voice warning function to remind users to charge in time. At the same time, recommended charging stations are marked on the navigation map, and the estimated remaining power upon arrival at each station, as well as the evaluation parameter matrix of each station, are displayed to help users fully understand the situation of surrounding charging stations and make the best charging decisions.

[0046] A multi-source data fusion dynamic optimization system and management method for electric vehicle endurance, the method comprising the following steps:

[0047] Step S100: collecting multi-source data of the electric vehicle, including vehicle parameter data, wind speed data, and driving behavior data, and pre-processing the data;

[0048] Step S200: Based on the pre-processed multi-source data, predict the endurance of the electric vehicle and predict the remaining endurance distance;

[0049] Step S300: evaluating the charging stations on the planned route of the electric vehicle based on the prediction results, and generating a recommended sequence of charging stations based on the evaluation results;

[0050] Step S400: providing feedback to the user based on the charging station recommendation sequence.

[0051] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: the electric vehicle endurance dynamic optimization system and management method with multi-source data fusion provided by the present invention include a vehicle data acquisition module, an energy consumption prediction module, a dynamic optimization module and a feedback module: the vehicle data acquisition module is used to collect multi-source data of the electric vehicle and perform preprocessing; the energy consumption prediction module is used to predict the endurance of the electric vehicle and the remaining endurance distance based on the multi-source data acquired by the vehicle data acquisition module; the dynamic optimization module includes an electricity price prediction unit, a queuing prediction unit and an evaluation unit, which are used to evaluate the charging stations on the planned route of the electric vehicle according to the prediction results of the energy consumption prediction module, and generate a charging station recommendation sequence according to the evaluation results; the feedback module is used to provide feedback to the user based on the charging station recommendation sequence of the dynamic optimization module; the present invention improves the accuracy of endurance calculation by combining actual terrain; further improves the practicality of the charging recommendation plan by introducing electricity prices and queuing time costs, thereby improving user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0053] Figure 1 It is a structural diagram of an embodiment of the present invention;

[0054] Figure 2 Schematic diagram of the steps of an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] See also Figure 1-Figure 2, providing a multi-source data fusion electric vehicle endurance dynamic optimization system, which includes a vehicle data acquisition module, an energy consumption prediction module, a dynamic optimization module and a feedback module:

[0057] The vehicle data acquisition module is used to collect multi-source data of the electric vehicle, including vehicle parameter data, wind speed data and driving behavior data, and perform pre-processing;

[0058] The energy consumption prediction module is used to predict the cruising range of the electric vehicle and the remaining cruising distance based on the multi-source data acquired by the vehicle data acquisition module;

[0059] The dynamic optimization module includes an electricity price prediction unit, a queue prediction unit, and an evaluation unit, which is used to evaluate the charging stations on the planned route of the electric vehicle based on the prediction results of the energy consumption prediction module and generate a recommended sequence of charging stations based on the evaluation results;

[0060] The feedback module is used to provide feedback to the user based on the charging station recommendation sequence of the dynamic optimization module.

[0061] Specifically, the vehicle data acquisition module includes a data acquisition unit and a data preprocessing unit:

[0062] The data acquisition unit connects to the map software through the API based on the real-time GPS coordinates of the vehicle and parses the road topology data returned by the map service. In combination with the on-board sensors, the on-board network and the external environment monitoring equipment, the data acquisition unit obtains the vehicle parameter data of the electric vehicle in real time. The vehicle parameter data includes battery parameters and operating parameters. The battery parameters include battery state of charge, battery temperature, battery internal resistance and battery health status; the operating parameters include vehicle speed, acceleration, mileage and road gradient. The data acquisition unit connects to the national meteorological data platform through the HTTPS protocol, generates a geographic grid code based on the vehicle location, requests real-time meteorological observation data within a preset range, extracts the wind speed vector and obtains the wind speed data. The data acquisition unit directly reads the original signal of the on-board diagnostic system through the CAN bus monitor to obtain driving behavior data in real time.

[0063] The data preprocessing unit is used to synchronize the vehicle parameter data, wind speed data and driving behavior data collected by the data acquisition unit; perform timestamp alignment operations through the GPS clock and PTP protocol, and periodically calibrate the vehicle parameter data, wind speed data and driving behavior data.

[0064] Through on-board sensors, on-board networks, and external environmental monitoring equipment, multi-dimensional data is acquired in real time during vehicle operation. Vehicle battery-related data includes battery state of charge (SOC), battery temperature, battery internal resistance, and battery state of health (SOH). These parameters directly reflect the battery's performance and remaining power. Vehicle operating parameters include speed, acceleration, mileage, road gradient, and other parameters that reflect the vehicle's driving status and operating conditions. Environmental data includes ambient temperature and humidity, wind speed and direction, and other environmental factors that have a significant impact on vehicle energy consumption. Furthermore, the system also collects driving behavior data, such as rapid acceleration and deceleration frequency, and average speed changes. These data can reflect the driver's driving habits and thus affect vehicle energy consumption.

[0065] Because the collected data may contain noise, outliers, or inconsistent formats, the preprocessing unit cleans and standardizes the raw data. Using algorithms such as data filtering and outlier detection and correction, it removes noise and erroneous information from the data. It also converts data of different types and formats into a unified format for analysis and processing by subsequent modules.

[0066] Specifically, the energy consumption prediction module includes:

[0067] A dynamic energy consumption model is constructed using a long short-term memory (LSTM) neural network. The input layer receives preprocessed vehicle parameter data, wind speed data, and driving behavior data in a time series format. Each time step corresponds to a set of data containing battery parameters, operating parameters, wind speed, and driving behavior parameters. Multiple LSTM layers are set up, and a fully connected layer integrates the feature vectors output by the LSTM layers. The output layer calculates and outputs the vehicle's remaining range based on the results of the fully connected layer.

[0068] Extract the working log of the electric vehicle, extract historical vehicle parameter data, wind speed data, driving behavior data and the remaining range of the vehicle, and divide the historical data into training set and validation set according to the specified ratio;

[0069] The dynamic energy consumption model is trained using the training set, and the mean square error is used as the loss function to measure the error between the remaining cruising distance predicted by the model and the actual remaining cruising distance. Based on the validation set, the model parameters are iteratively updated using stochastic gradient descent and its improved algorithm, and the weights and biases in the model are continuously adjusted until the loss function value is less than or equal to a preset loss threshold. The iterative operation is stopped to obtain a trained dynamic energy consumption model.

[0070] The preprocessed real-time multi-source data is input into the trained dynamic energy consumption model in sequence according to time steps, and the predicted value of the vehicle's remaining cruising distance in the current state is output in the output layer.

[0071] The energy consumption prediction module is one of the system's core modules. Its primary function is to accurately predict the vehicle's remaining range based on preprocessed data provided by the vehicle data acquisition module. This module uses a long short-term memory (LSTM) neural network to construct a dynamic energy consumption model. LSTM neural networks have powerful time series data processing capabilities, effectively capturing the dynamic impact of various factors on energy consumption during vehicle operation. The model inputs preprocessed data such as battery SOC, temperature, vehicle speed, acceleration, ambient temperature and humidity, and road gradient. After learning and calculating, the LSTM neural network outputs the vehicle's remaining range. To ensure accurate and timely predictions, the system updates the predicted value every five minutes. Furthermore, it dynamically adjusts the energy consumption coefficient based on real-time traffic information, such as traffic congestion and changes in road gradient. For example, on congested roads, where vehicles frequently start and stop, energy consumption can increase significantly. The system promptly adjusts the energy consumption coefficient based on real-time traffic data to ensure that the remaining range prediction is more accurate.

[0072] Specifically, the energy consumption prediction module further includes updating the remaining endurance prediction value once every time node.

[0073] Specifically, the electricity price prediction unit includes:

[0074] Based on the planned path of the electric vehicle and the remaining cruising distance, obtaining a maximum cruising path of the electric vehicle;

[0075] Based on the maximum endurance path, all charging stations on the maximum endurance path are obtained to construct a charging station set;

[0076] For charging station A in the charging station set i To make electricity price forecasts:

[0077] By analyzing the historical load data, power generation cost data, and power market transaction data of the power grid along the planned path and combining it with a time series prediction algorithm, the power price trend of the power grid in the future period is predicted.

[0078] Based on charging station A i The control coefficient of the charging station is calculated based on the historical electricity price data and the grid electricity price data of the corresponding period;

[0079] Based on the electricity price trend of the power grid in the future period, the electricity price trend of the charging station in the future period is obtained.

[0080] By analyzing historical grid load data, generation cost data, and electricity market transaction data, combined with a time series prediction algorithm, we can predict electricity price trends for the next two hours at each charging station along the planned route. For example, during peak hours, when the grid load is high, electricity prices may rise; while during low-demand hours at night, electricity prices are relatively low. By accurately predicting electricity price trends, users can optimize their charging schedules and reduce charging costs.

[0081] Specifically, the queue prediction unit includes:

[0082] Connect to charging station A i Real-time occupancy data, including the usage status of charging piles and the number of vehicles currently queuing, and statistics on the number of charging piles in working condition;

[0083] Based on the queuing theory model, input charging station A i Output the number of normal charging piles, the number of charging piles in working state, the number of vehicles in queue and the historical average service time of charging station A i The expected waiting time.

[0084] The system uses real-time charging station occupancy data, including the usage status of charging piles and the number of vehicles currently in the queue, to calculate the expected wait time at each charging station based on queuing theory models, such as the M / M / c queuing model. The M / M / c queuing model is a classic queuing theory model, where "M" represents an exponential distribution for both the arrival and service processes, and "c" represents the number of charging piles. By inputting parameters such as the number of charging piles, the current number of vehicles in the queue, and the historical average service time, the model can accurately predict user wait times at each charging station, helping users avoid peak queues and improve charging efficiency.

[0085] Specifically, the assessment units include:

[0086] According to charging station A i The location data and the average speed of the electric vehicle are used to calculate the speed at which the vehicle reaches charging station A. i duration;

[0087] Based on charging station A i The estimated waiting time of charging station A is calculated i The actual waiting time;

[0088] Construct a multi-objective evaluation function Score = α*(the electricity price at the time of the expected waiting time) + β*(the actual waiting time) + γ*(the charging station A i The distance from the maximum endurance path) for charging station A i Perform a comprehensive evaluation; Score represents the comprehensive score, α, β, and γ represent weights, and α, β, and γ support user-defined configuration;

[0089] Perform a comprehensive evaluation on all charging stations in the charging station set, calculate the comprehensive score of each charging station, and sort the charging stations from high to low according to the comprehensive score to generate a recommended charging station sequence.

[0090] Users can adjust the weights of each indicator according to their own needs, such as paying more attention to economic costs (increasing the α value) or paying more attention to time efficiency (increasing the β value). The system will calculate the comprehensive score of each charging station based on the weight set by the user and sort the charging stations according to the score. By inputting parameters such as the number of charging piles, the current number of vehicles in the queue, and the historical average service time, the model can accurately predict the user's waiting time at each charging station, helping users avoid queue peaks and improve charging efficiency.

[0091] Specifically, the feedback module includes:

[0092] When the vehicle's remaining range falls below the preset range threshold, the feedback module will automatically trigger a voice warning function to remind the user to charge in time; it will mark recommended charging stations on the navigation map, display the estimated remaining power upon arrival at each station, and a comprehensive scoring matrix for each station.

[0093] The feedback module is responsible for presenting the system's analysis results to users in an intuitive and convenient manner, supporting their charging decisions. Its main functions include:

[0094] Generation and display of charging station priority lists: Based on the comprehensive scores of charging stations calculated by the dynamic optimization module, a charging station priority list is generated, and the three best charging stations and their evaluation parameters, such as electricity price, estimated waiting time, distance, etc., are displayed to users through the vehicle head-up display (HUD) system. Users can quickly obtain key information without having to look down at their mobile phones or vehicle screens, thereby improving driving safety. Battery life warning and navigation: When the vehicle's remaining battery life is lower than the preset threshold, the feedback module will automatically trigger the voice warning function to remind users to charge in time. At the same time, recommended charging stations are marked on the navigation map, and the estimated remaining power upon arrival at each station, as well as the evaluation parameter matrix of each station, are displayed to help users fully understand the situation of surrounding charging stations and make the best charging decisions.

[0095] A multi-source data fusion dynamic optimization system and management method for electric vehicle endurance, the method comprising the following steps:

[0096] Step S100: collecting multi-source data of the electric vehicle, including vehicle parameter data, wind speed data, and driving behavior data, and pre-processing the data;

[0097] Step S200: Based on the pre-processed multi-source data, predict the endurance of the electric vehicle and predict the remaining endurance distance;

[0098] Step S300: evaluating the charging stations on the planned route of the electric vehicle based on the prediction results, and generating a recommended sequence of charging stations based on the evaluation results;

[0099] Step S400: providing feedback to the user based on the charging station recommendation sequence.

[0100] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0101] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. The electric vehicle endurance dynamic optimization system based on multi-source data fusion is characterized by: The system includes a vehicle data acquisition module, an energy consumption prediction module, a dynamic optimization module, and a feedback module: The vehicle data acquisition module is used to collect multi-source data of the electric vehicle, including vehicle parameter data, wind speed data and driving behavior data, and perform pre-processing; The energy consumption prediction module is used to predict the cruising range of the electric vehicle and the remaining cruising distance based on the multi-source data acquired by the vehicle data acquisition module; The dynamic optimization module includes an electricity price prediction unit, a queue prediction unit, and an evaluation unit, which is used to evaluate the charging stations on the planned route of the electric vehicle based on the prediction results of the energy consumption prediction module and generate a recommended sequence of charging stations based on the evaluation results; The feedback module is used to provide feedback to the user based on the charging station recommendation sequence of the dynamic optimization module.

2. The electric vehicle endurance dynamic optimization system based on multi-source data fusion according to claim 1 is characterized in that: The vehicle data acquisition module includes a data acquisition unit and a data preprocessing unit: The data acquisition unit connects to the map software through the API according to the real-time GPS coordinates of the vehicle and parses the road topology data returned by the map service. In combination with the on-board sensors, the on-board network and the external environment monitoring equipment, the data acquisition unit obtains the vehicle parameter data of the electric vehicle in real time. The vehicle parameter data includes battery parameters and operating parameters. The battery parameters include battery state of charge, battery temperature, battery internal resistance and battery health status; the operating parameters include vehicle speed, acceleration, mileage and road gradient. Connect to the national meteorological data platform via the HTTPS protocol, generate a geographic grid code based on the vehicle's location, request real-time meteorological observation data within a preset range, extract the wind speed vector, and obtain wind speed data; Directly read the original signal of the on-board diagnostic system through the CAN bus monitor to obtain driving behavior data in real time; The data preprocessing unit is used to synchronize the vehicle parameter data, wind speed data and driving behavior data collected by the data acquisition unit; perform timestamp alignment operations through the GPS clock and PTP protocol, and periodically calibrate the vehicle parameter data, wind speed data and driving behavior data.

3. The electric vehicle endurance dynamic optimization system based on multi-source data fusion according to claim 2 is characterized in that: The energy consumption prediction module includes: A dynamic energy consumption model is constructed using a long short-term memory (LSTM) neural network. The input layer receives preprocessed vehicle parameter data, wind speed data, and driving behavior data in a time series format. Each time step corresponds to a set of data containing battery parameters, operating parameters, wind speed, and driving behavior parameters. Multiple LSTM layers are set up, and a fully connected layer integrates the feature vectors output by the LSTM layers. The output layer calculates and outputs the vehicle's remaining range based on the results of the fully connected layer. Extract the working log of the electric vehicle, extract historical vehicle parameter data, wind speed data, driving behavior data and the remaining range of the vehicle, and divide the historical data into training set and validation set according to the specified ratio; The dynamic energy consumption model is trained using the training set, and the mean square error is used as the loss function to measure the error between the remaining cruising distance predicted by the model and the actual remaining cruising distance. Based on the validation set, the model parameters are iteratively updated using stochastic gradient descent and its improved algorithm, and the weights and biases in the model are continuously adjusted until the loss function value is less than or equal to a preset loss threshold. The iterative operation is stopped to obtain a trained dynamic energy consumption model. The preprocessed real-time multi-source data is input into the trained dynamic energy consumption model in sequence according to time steps, and the predicted value of the vehicle's remaining cruising distance in the current state is output in the output layer.

4. The electric vehicle endurance dynamic optimization system based on multi-source data fusion according to claim 3 is characterized in that: The energy consumption prediction module also includes updating the remaining endurance prediction value once every time node.

5. The electric vehicle endurance dynamic optimization system based on multi-source data fusion according to claim 1 is characterized in that: The electricity price prediction unit includes: Based on the planned path of the electric vehicle and the remaining cruising distance, obtaining a maximum cruising path of the electric vehicle; Based on the maximum endurance path, all charging stations on the maximum endurance path are obtained to construct a charging station set; For charging station A in the charging station set i To make electricity price forecasts: By analyzing the historical load data, power generation cost data, and power market transaction data of the power grid along the planned path and combining it with a time series prediction algorithm, the power price trend of the power grid in the future period is predicted. Based on charging station A i The control coefficient of the charging station is calculated based on the historical electricity price data and the grid electricity price data of the corresponding period; Based on the electricity price trend of the power grid in the future period, the electricity price trend of the charging station in the future period is obtained.

6. The electric vehicle endurance dynamic optimization system based on multi-source data fusion according to claim 5 is characterized in that: The queue prediction unit includes: Connect to charging station A i Real-time occupancy data, including the usage status of charging piles and the number of vehicles currently queuing, and statistics on the number of charging piles in working condition; Based on the queuing theory model, input charging station A i Output the number of normal charging piles, the number of charging piles in working state, the number of vehicles in queue and the historical average service time of charging station A i The expected waiting time.

7. The electric vehicle endurance dynamic optimization system based on multi-source data fusion according to claim 6 is characterized in that: Assessment units include: According to charging station A i The location data and the average speed of the electric vehicle are used to calculate the speed at which the vehicle reaches charging station A. i duration; Based on charging station A i The estimated waiting time of charging station A is calculated i The actual waiting time; Construct a multi-objective evaluation function to evaluate charging station A. i Conduct comprehensive assessments; Perform a comprehensive evaluation on all charging stations in the charging station set, calculate the comprehensive score of each charging station, and sort the charging stations from high to low according to the comprehensive score to generate a recommended charging station sequence.

8. The electric vehicle endurance dynamic optimization system based on multi-source data fusion according to claim 1 is characterized in that: The feedback module includes: When the vehicle's remaining range falls below the preset range threshold, the feedback module will automatically trigger a voice warning function to remind the user to charge in time; it will mark recommended charging stations on the navigation map, display the estimated remaining power upon arrival at each station, and a comprehensive scoring matrix for each station.

9. The electric vehicle endurance dynamic optimization system and management method based on multi-source data fusion is characterized by: The method comprises the following steps: Step S100: collecting multi-source data of the electric vehicle, including vehicle parameter data, wind speed data, and driving behavior data, and pre-processing the data; Step S200: Based on the pre-processed multi-source data, predict the endurance of the electric vehicle and predict the remaining endurance distance; Step S300: evaluating the charging stations on the planned route of the electric vehicle based on the prediction results, and generating a recommended sequence of charging stations based on the evaluation results; Step S400: providing feedback to the user based on the charging station recommendation sequence.

Images