在对路段安全状态和山区高速公路进行定义的前提下，提出了山区高速公路路段的定义及其安全状态评价指标及评价标准．通过在需要进行评价的路段以及路段的上游安装交通流信息采集设备来获取速度参数．以山区高速公路交通事故历史数据和实时的速度参数作为评价指标，参考提出的评价标准，建立了基于概率神经网络( PNN) 的路段安全状态评价模型．仿真结果表明，该模型能够对山区高速公路路段安全状态进行评价．

In this paper,first,the concepts of safety state of section and the freeway in mountains are defined and under such definitions,the concept of freeway section in mountains as well as the safety evaluation indexes and standard of the section is put forward. Then,the speed parameters are obtained by installing the traffic information collection equipment near the section requiring evaluation and its upstream. Finally,with real －time speed parameters and historical data of traffic accidents as the evaluation indexes,a safety state evaluation model of freeway section in mountains is established based on the proposed evaluation standard and the probabilistic neural network ( PNN) . Simulated results indicate that the proposed model is capable of evaluating the safety state of freeway section in mountains.