Research On Vehicle Speed Prediction Algorithm Based On Driver-Vehicle-Road-Traffic System

Faced with an important turning point in the upcoming energy structure,the new energy automobile industry has also set off a new wave of technological revolution.However,traditional new energy vehicles still rely on existing typical road conditions in terms of energy management systems.These typical road conditions are universal,but different vehicles will produce completely different working conditions data for different drivers in different traffic environments.It will greatly reduce the effect of energy saving and emission reduction of new energy vehicles.And for drivers driving new energy vehicles,the unknown of the future driving environment also makes the driver unable to develop a more sensible driving strategy,spending more effort and energy on the road.Therefore,the above problem can be solved by accurately predicting the speed of the vehicle in the future driving conditions and applying it to the intelligent energy management system and the intelligent navigation system of the new energy vehicle.So,faced with the limitations of traditional and current vehicle speed prediction algorithms,this paper proposes the vehicle speed of short-term and long-term prediction algorithm,based on human-vehicle-road-traffic system and artificial intelligence algorithms,which are used in three scenarios including power prediction,driving energy consumption prediction,and driving time consumption prediction.The details are as follows:(1)Through the analysis of human-vehicle driving characteristics,road and traffic flow,the factors affecting the vehicle speed prediction are determined,and the data is collected by design experiments.Then,the Pearson correlation coefficient method is used to extract key features of short-term vehicle speed prediction model and the long-term vehicle speed prediction model,thereby reducing the model training burden and improving computational efficiency.(2)For short-term vehicle speed prediction,based on the extracted key features,a variety of artificial intelligence algorithms are used to predict and analyze the short-term vehicle speed,and the optimal algorithm model is built,which is the optimizes GA-SVM model.And the short-term vehicle speed prediction model is finally constructed.The short-term vehicle speed prediction result is applied to the power prediction to evaluate its prediction effect.(3)For long-term vehicle speed prediction,based on the extracted key features,a variety of typical artificial intelligence algorithms are used to predict and analyze the long-term vehicle speed,and the optimal algorithm model is built,which is the BP-LSTM model.And the long-term vehicle speed prediction model is finally constructed.The long-term vehicle speed prediction result is applied to vehicle driving energy consumption prediction and vehicle travel time prediction to evaluate their prediction effects.The research results show that the optimized GA-SVM prediction model and BP-LSTM prediction model can achieve better prediction results in short-term vehicle speed prediction and long-term vehicle speed prediction respectively,and have strong practicability,which has a certain reference value for energy management and driving strategy formulation.