论文部分内容阅读
普通公路安全预警尤其是山区公路安全预警,由于其交通流的复杂性使得利用监控系统检测交通状态较为困难。离散选择分析模型通过原始数据直接做出判断,又可以根据数据更新系统进行有效的自身学习,因此较目前应用的贝叶斯网络模型更适用于山区公路安全预警。首先利用离散选择分析方法建立了道路交通状态检测的判断模型;其次以检测结果为基础,利用线性神经网络模型建立了路段交通预警模型;最终建立了山区公路安全预警系统。
Early Warning of Common Highway Safety, especially Warning of Highway Safety in Mountainous Areas, makes it more difficult to monitor the traffic status by using the monitoring system because of the complexity of traffic flow. Discrete choice analysis model can judge directly from the original data and can effectively learn itself according to the data update system. Therefore, the Bayesian network model is more suitable for early warning of mountain highway safety. Firstly, the judgment model of road traffic condition detection is established by using the method of discrete choice analysis. Secondly, based on the test results, the traffic warning model of road section is established by linear neural network model. Finally, the warning system of mountain highway safety is established.