中文摘要：

为了提高高速公路交通事故检测效率并降低误警率,提出了一种基于小波方程的高速公路交通事故自动检测方法。定义了能够反映占有率和速度累积变化的小波方程,构建了事故检测统计量集合,利用突变值确定了最优检测统计量。考虑到检测阈值在不同交通运行状态下对检测效率的影响,通过引入误警贡献度的概念,建立了期望误警率、检测统计量标准差、误警贡献度间的函数关系,提出了一种检测阈值自适应变化策略。研究结果表明：基于小波方程的事故检测算法具有较高的检测率和较低的误警率、以及更快的平均检测时间,在综合各种交通运行情况下,平均检测率可达到94%,误警率可降至1%;自适应阈值策略对事故检测效率影响显著,对交通运行环境的变化具有较强的适应性。

英文摘要：

In order to improve the detection rate and lower the false alarm rate, an automatic expresway traffic accident detection algorithm based on wavelet equation was proposed. Wavelet equations were defined to illustrate the cumulative changes of occupancy and speed, with which the statistic set for accident detection was constructed. The best test statistic was selected by the measurement of divergence. With the consideration of the influence of detection threshold on detection efficiency under different traffic conditions, an adaptive change strategy for detection threshold was proposed. To implement the adaptive change strategy for detection threshold, the functional relations among desirable false alarm rate, standard deviation of detection statistics and false alarm contribution were established. The results show that the proposed algorithm based on wavelet equation has a higher average detection rate （94%）, lower average false alarm rate （1%） and shorter average detection time. The adaptive threshold strategy has significant influence on detection efficiency and strong adaptability under various traffic conditions.