Economic Vehicle Speed Optimization Method Based On Traffic Condition Prediction

Energy security and environmental protection put the automotive industry under the tremendous pressure to energy saving and emission reduction.When a vehicle passes through a continuous signal intersection section,the excessive or slow speed of the vehicle may generate a large number of acceleration and idling processes,which may increase fuel consumption.An economic speed optimization method considering traffic state is proposed based on those conditions.When a vehicle obtains the traffic information in advance,the dynamic economic speed of the vehicle is calculated by the economic speed optimization method considering the traffic state,which ensures the timeliness of driving,energy conservation and emission reduction.The main research content and conclusions are as follows:(1)A fuel consumption model on vehicle evaluation is established.The required type of fuel consumption model is determined through analysis.The basis of modeling is the public fuel consumption data provided by ANL of the United States.Through data filtering and smoothing processing,the relationship between power and fuel consumption is drawn and the fuel consumption model is obtained by fitting.(2)A traffic state prediction method based on deep belief network is proposed to predict the feasible speed range required by the economic speed optimization system.A single section traffic state prediction method is proposed,and the DBN model is used as prediction model.The single-section traffic state prediction method is extended to a multi-section traffic state prediction method based on rolling prediction.It is found that DBN model has a better prediction performance compared with several common traffic flow prediction models.The phase of signal lights has an effect on the prediction process,and the vehicle data at the intersection at green time has a better effect on the prediction accuracy of the model than the red time.(3)The economic speed optimization method of continuous signal intersection section based on traffic state construction is proposed.The economic speed optimization foundation of single road section and intersection is constructed respectively,and the economic speed of vehicles at continuous intersections is optimized by using the multi-section traffic state prediction method based on rolling prediction.By constructing d-t plane and discretizing time nodes,Dijkstra algorithm is used to find a path with the lowest fuel consumption among many speed combinations.In order to verify the superiority of the method,three driving strategies are ana-lyzed through examples.The analysis found that the economic speed optimization method considering traffic conditions has a good performance on fuel saving and travel time efficiency,so that the driver has a better driving experience.