在道路交通和城市化进程影响下对汽车销量的预测

本文研究了汽车销量在道路交通和城市化进程影响下的预测模型的建立,以历史数据为检测依据,判断所建模型的可行性,并对2011~2015年末汽车总销量进行了预测。其模型参数即道路交通和城市化进程对汽车销量的影响因素包括城镇人口数、城镇居民人均可支配收入、城市化率、城市年末实有道路长度、城市年末实有道路面积、公路里程。经精度检验,BP神经网络可用于预测城镇人口数、城镇居民人均可支配收入、城市化率、公路里程。二元线性回归可用于预测城市道路长度和城市道路面积。最后BP神经网络利用其在处理非线性系统方面的优越性,实现在各参数影响输入下对汽车总销量输出的任意非线性映射且具有较高的预测精度。

Forecasting Vehicle Quantity under the Influence of Road Traffic and the Process of Urbanization

This paper focus on setting forecasting model of car quantity under the influence of road traffic and the process of urbanization which is judged whether it possesses feasibility on inspection basis of historical statistics,and forecasting car quantity from 2011 to 2015.Model factor which is exposed to road traffic and the process of urbanization includes urban population,urban per capita disposable income,urbanization rate,urban annual practical length of road,urban annual practical area of road and highway mileage.By precision examination,BP nerve net is available to forecast urban population,urban per capita disposable income,urbanization rate and highway mileage.Binary linear regression is available to urban annual practical length of road and urban annual practical area of road.At last,BP nerve net uses the priority on handling nonlinear system and brings about arbitrary nonlinear mapping of car quantity from 2011 to 2015 under entering the influence of all factors.