基于车流量的信号灯及车道数智能调节系统设计

Design of Intelligent Regulation System for Traffic Lights and Lane Numbers Based on Traffic Volume

针对城市道路交通流的不确定性和随机波动性,应用基于BP神经网络算法和韦伯斯特模型来优化城市道路,提高车辆通行效率。利用神经网络算法较高的预测精确度对交通流进行预测,并以此预测结果为基础,提前对导向可变车道进行变换,由此提高了道路的空间利用率;然后依据车道属性改变后的数据,运用改进后的韦伯斯特模型计算信号灯配时参数,并结合当前道路的具体状况来优化红绿灯配时,由此提高了路口的时间利用率。以合肥市某交叉路口交通流量数据对提出的方法进行测试,验证了该方法的有效性。

With the uncertainty and fluctuation of the traffic flow,a BP neural network algorithm and Webster model are applied to optimize urban roads and improve traffic effectiveness.In this method,the traffic volume is forecast with the high prediction accuracy of the neural network algorithm,and the variable direction lanes are switched based on the forecast data,which can improve the utilization ratio of the roads.According to the data of the changed lane attributes,the signal timing parameters are calculated through the improved Webster model and the signal timing dial is optimized with the specific conditions of the current road.In this way,the time utilization rate at the intersection is improved.With the traffic volume data at an intersection in Hefei as an example,this proposed method has been tested and it proves to be effective.