һ�ָ��Ƶ�·��ֲڶ����۳�������

鉴于对国家公路系统进行全面评价的方法耗费太大，一般采用抽样方法来评价路面粗糙度，这涉及到很多实际因素．尽管分层随机抽样的方法已经在干线和一般的集散系统得到利用，它仍有很大的发展空间．本文在提出多个针对路面条件的随机变量的基础上，通过改变路面粗糙度分布的基本假定来改进抽样方法，为路面粗糙度的估计建立了一个整体的框架．论文简要介绍了简单随机抽样、分层随机抽样方式，通过分析说明后者能够对路面网络的粗糙度样本提供更全面的估计，而且其偏差更小．论文进一步讨论了以精确估计为基础的良好分层带来的影响．结果表明，在交通网络上可找到一种独特的优化分层方法．根据分析结果，本文定义了该改进分层方法．

Improved Sampling Method for Pavement Roughness Evaluation

The sampling method for pavement roughness evaluation has significant practical implications because the exhaustive review method in use for Nation Highway System (NHS) is too costly. Though stratified random sampling method is adopted for the arterial and collector functional systems, there is a good potential to improve it.In this paper, we build a general framework for pavement roughness evaluation by improving the sampling method with fundamental assumptions about the distribution of pavement roughness, based on which we define the random variable for the overall pavement condition. After briefly introducing a simple random sampling and a stratified random sampling method, we analytically show that the latter can provide an estimate of the comprehensive roughness profile over the network with a smaller deviation. We further discuss the impact of a good stratification on an accurate estimation. Analytically we show that there exists a unique optimal stratification method applied to the transportation network. Based on the analytical result, we propose an improved stratification method.