摘要
交通需求预测是道路建设项目规划工作的重要组成部分,是确定道路等级、技术标准及进行规划决策、投资效益分析的依据。居民出行发生量预测是交通需求预测的一个重要组成部分,通常使用Gompertz模型,但其模型系数一直是通过类比估算方法确定,缺乏依据。本研究利用最小二乘法改进待定参数估计,推算Gompertz模型系数确定公式,采用Matlab程序分析得到Gompertz模型的系数。利用残差估计和类比发现,使用最小二乘法拟合得到的Gompertz模型系数具有明显的优点。根据山东省东营市交通调查数据,利用本研究改进模型,预测该市2025年的居民出行量。结果表明,使用改进Gompertz模型系数求解方法能有效提高居民出行发生量的预测精度,对合理规划道路网结构具有一定指导意义。
The predictions of traffic and transportation demands are important and serve as basis in determining the road level, formulating technical standards and making plans as well as evaluating investment. A tripgeneration forecast is an important part of the demand forecast, which is generally acquired based on Gompertz model. However, the parameters of Gompertz model estimated by analogy are lack of a sound basis. Therefore, this paper proposes to improve the estimation efficiency of undetermined parameters in Gompertz model with the least-squares method and obtains more reliable model parameters by Matlab. Its better performance is shown by the residual estimation and analogy. Statistical data from Dongying, Shandong Provinces, is analyzed by the proposed model to predict the trip generation in 2025. The final result shows that this proposed Gompertz model can improve effectively the prediction accuracy and may be used in rational road network planning.
出处
《科技导报》
CAS
CSCD
北大核心
2009年第18期82-85,共4页
Science & Technology Review
基金
山东交通学院科研基金项目(Z200907)
