基于支持向量机的高速公路事件检测研究

Expressway Incident Detection Using Support Vector Machines

高速公路事件检测(A ID)系统可以有效的减少交通延误、保障道路安全,减少环境污染,是智能交通管理系统的重要组件。文章采用一种新的模式识别技术——支持向量机(SVM),来进行高速公路的事件检测。支持向量机是以统计学习理论的VC维理论和结构风险最小化原理为基础,算法上采用二次规划、拉格朗日理论,因此有效避免了过学习、局部极小点等问题。最后,采用M ATLAB对该方法进行了测试。

Expressway incident detection （AID） system can effectively reduce traffic delays, ensure road safety and protect environment, which is an important part of Advanced Traffic Management and Information System （ATMIS）. This paper presents the application of a recentlydeveloped pattern classifier called support vector machine （SVM） in expressway incident detection. Support vector machines （SVM）, based on VC-dimension theory and the structural risk minimization principle, implement quadratic programming and Lagrange method. Thus, overfitting and local optimal solution are unlikely to occur with SVM. At last, Matlab is applied to check the new method.