基于BP神经网络的北京市出租车油耗模型研究

Study on Taxi Fuel Consumption Model in Beijing Based on BP Neural Network

从宏观的角度出发,在传统油耗算法研究的基础上,从影响油耗的道路和交通特性方面,确定了3个输入参数,分别是链路平均速度、交叉口密度和停驶比,并利用大量出租车油耗数据分析了北京市主干路上这3个影响因素与油耗之间的关系,提出了基于BP神经网络的油耗模型,结果表明神经网络具有更高的精度和稳定性,针对北京市路网,选择具有代表性的长安街和西大望路进行了实例应用.研究表明:对城市道路进行信号联动控制或对车辆进行路径诱导、减少停驶比、提高路段平均速度是降低能耗的一个有效方法.

Based on the research of traditional fuel consumption model,this paper identified three main factors in terms of road and traffic conditions,namely the average speed of the link,the intersection density,and stop ratio. And it analyzed the relationship between the three influence factors and fuel consumption on the main road in Beijing on the base of a large number of the taxi fuel consumption data.Then a fuel consumption model is proposed based on the BP neural network. The results show that the neural network model has a higher accuracy and stability. In view of the Beijing city road network,it applied the representative example of Chang＇an Avenue and West Da Wang Road. Research results show that it is an effective method to reduce the energy consumption by using coordinated signal control or dynamic route guidance.