基于改进BP神经网络的道路交通事故预测

道路交通事故因受多种随机因素的影响而呈现出非线性的特点,传统的线性分析方法无法完全揭示其内涵。在分析道路交通事故与人、车、路等因素关系的基础上,利用神经网络具有描述非线性特性的能力,将影响交通事故的多种因素综合起来建立了基于改进BP神经网络的道路交通事故预测模型。选取人口密度、路网密度和机动车辆密度作为交通事故预测模型的输入神经元,采用道路交通综合死亡率作为道路交通事故的输出评价指标,对道路交通事故进行预测。实验结果表明,该预测模型能很好地适用于道路交通事故预测,验证了该模型的可行性和有效性。

Forecasting Model of Road Traffic Accident Based on Improved BP Neural Network

Due to the fact that it is affected by various random factors, a traffic accident is nonlinear in nature. Thus, its essence can not be efficiently revealed by traditional linear analysis method. Starting from the analysis of the relation between traffic accident and factors, including human, vehicle and road, and employing, the nonlinear characteristics described by a neural network of a road traffic accident, a forecasting model based on improved BP, is proposed by integrating the factors affecting traffic. A traffic accident prediction model uses population density,road network density, and motor vehicle density as the input neurons and an output neuron which is road accident comprehensive mortality. The results show that the improved BP neural network is well-suited for the forecasting of road traffic accidents, thus, verifing the feasibility and effectiveness of the model is verified.