The Prediction Of Travel Energy Consumption Of Electric Vehicle With Traffic Information

The development of electric vehicles is an important way to solve the problems of environmental pollution and depletion of oil resources at this stage.In the process of using electric vehicles,mileage anxiety has become one of the most important factors that affect the driving experience due to the limitation of battery capacity.Robust and accurate prediction of the energy consumption of the journey of the electric vehicle can guide the driver to allocate the power rationally and relieve the anxiety of the mileage.However,the current energy consumption forecasting method of electric vehicles does not consider the influence of traffic conditions on the actual working conditions during the driving process,which leads to the low prediction accuracy and cannot effectively alleviate the driver’s mileage anxiety.Considering the actual working conditions of vehicles,this paper integrates the traffic information of electric vehicles and uses depth learning method to make the energy consumption prediction of electric vehicles.A prediction method of travel energy consumption of electric vehicles based on deep learning model combined with attention mechanism is proposed.The simulation results show that compared with the previous methods,the prediction model of travel energy consumption proposed in this paper can achieve higher prediction accuracy.The main work and research results are as follows:（1）An electric vehicle running data acquisition and trip energy consumption simulation platform is been constructed.From which,a vehicle driving state and travel energy consumption simulation data set was constructed,and the problem of insufficient training data and insufficient training of travel energy prediction model was solved.Based on the actual running data of traffic network and electric vehicles,the vehicle driving conditions and the related parameters of vehicle’s important components are statistically analyzed.Traffic simulation model and pure electric vehicle simulation model are respectively constructed in Paramics traffic simulation software and Simulink / MATLAB environment.Real vehicle operating data show that the traffic network model and vehicle simulation model established in this paper have a high accuracy.The simulation model provides the data foundation for the subsequent training of prediction model.（2）This paper presents a method of traffic state feature extraction based on attention mechanism,which can extract effective key features from high-dimensional traffic network state data and reduce the interference of irrelevant features on travel energy consumption prediction model.According to the driving task,this method uses traffic data of the road network to calculate the corresponding attention weight of each road section of the road network during the driving task,so that the weighted state characteristics of the traffic network can better respresent the traffic status on the future travel energy consumption of the electric vehicle.The comparison test results show that adding the attention mechanism makes the mean absolute percent error of travel energy prediction model reduce by 2.95%.Combining the prediction results to visualize the attentional weights can visually show the impact of traffic status on energy consumption of future trips.The visualization results show that the method of traffic network state feature extraction based on attention mechanism is in line with human intuition.（3）A deep learning prediction method for energy consumption of electric vehicles based on traffic information is proposed.The method integrates vehicle status information and traffic network status features.In the Python environment,training and accuracy verification of energy consumption prediction model of electric vehicle based on traffic information are completed.The experimental results show that the average absolute error of model prediction is 0.0620 k W·h,and the average absolute percentage error is 8.22%.