Research On Motor Vehicle Ownership Warning Model Of Large And Medium-sized Cities Based On BP Neural Network

In the near decade, Chinese economy forwards steadily. GDP annual growth rate reach10.7%. It greatly promotes the process of urbanization in our country. It also has contributedto the great progress of motor vehicle industry. Because of the finite urban road area, parksand other infrastructure cannot satisfy the rapid growth of motor vehicles. It causes trafficproblems of urban to become increasingly serious. The sharp increase of motor vehicleownership which caused traffic safety problems such as environmental pollution caused theattention of people. How to scientifically forecast urban vehicle ownership and reasonablylimit the scale of the urban vehicle ownership have become a traffic management andimportant subject of urban planning department. This article established an early warningmodel of motor vehicle ownership based on the BP neural network, consumption of time andspace, and time series analysis. These models can feedback the realization of information inadvance. It lays a good foundation for layout and risk prevention timely. It also provides atheoretical basis and model to formulate policy for the relevant departments of government. Inthis paper, the main research work is as follows:(1) Established prediction model based on BP neural network. The growth of the motorvehicle ownership is under the influence of many factors. This article has chosen10impactindex including economic, population, and road traffic. It adopted a method of rollingforecasts and used the index of2001－2008to train the motor vehicle ownership of2002－2009. It uses the index of2002－2009to test motor vehicle ownership of2003－2010.Finally using2003－2010index to predict2011motor vehicle ownership.(2) It established the model of road network traffic capacity based on consumption oftime and space. It uses the time series of an exponential smoothing to forecast road net area.This article calculated the road network capacity of2011according to the prediction of roadnet area. According to the prediction of vehicle ownership and road network capacity we gotcoefficient of early warning.(3) Established the early warning model of vehicle ownership, including industry, thecoefficient, the description and judgment. The specific value of motor vehicle ownership androad network capacity can be as saturation of motor vehicle ownership. This model canpredict the size of the degree of saturation to city motor vehicle ownership in the coming year. So the government can make a reasonable planning on the future year according to the size ofthe future automobile ownership saturation.(4) It tests the model validation by using the example of Dalian. According to the actualdata we predict the motor vehicle ownership and road network traffic capacity of next twoyears. Their specific value could be the coefficient of motor vehicle ownership and analyzethat which level the industry is under and determine whether to need an alarm.Through the analysis of the instance, early warning model based on BP neural networkcan carry on the early warning to the motor vehicle ownership. It can predict warningcoefficient of motor vehicle ownership of the city future so as to realize feedback in advance.It is advantageous to the relevant department for possible traffic problems in advance.