基于LSTM-BP神经网络的道路拥堵时间智能预测

Intelligent Prediction of Road Congestion Time Based on LSTM-BP Neural Network

近年来,随着国民经济的发展,人民生活水平的提高,汽车占有率也在不断增加。然而道路建设的相对滞后,导致城市交通拥堵情况日益严重。现有的导航软件通过获取实时的GPS数据,确定当前的道路情况,但在严重的交通拥堵情况下,对交通拥堵时间预测的准确度较差。基于此,本文通过熵权法对模型的指标进行筛选,并提出一种LSTM-BP组合神经网络模型,该模型可以很好的识别车流量、平均旅行时间、平均速度等数据,以此提高模型的预测精度。

In recent years,with the development of the national economy and the improvement of people's living standards,the car share has also been increasing.However,the relatively lagging road construction has led to increasingly serious urban traffic congestion.Existing navigation software determines the current road conditions by acquiring real-time GPS data.However,in severe traffic congestion,the accuracy of this method for traffic congestion time prediction is poor.For this reason,this paper uses the entropy weight method to screen the model indicators,and proposes an LSTMBP combined neural network model,which can well identify traffic volume,average travel time,average speed and other data.Thereby improving the prediction accuracy of the model.