摘要
讨论了事件检测算法，建立了基于小波分析的算法。首先构造二进小波，对交通流数据进行二进小波变换，检测其奇异点，根据判断逻辑确定是否有事件发生。采用广佛高速公路交通流数据对该算法进行离线测试，与加利福尼亚算法、滤波算法比较，结果显示该算法具有较好的效果。
The paper discusses incident detection algorithms and puts forword a new method, Wavelet Algoritm. This new algoritm constructs a dynamic wavelet, makes dynamic wavelet transform on traffic flow data, detects abnormal signal, and according to the logic tree, draws conclusion. According to the result of an off-line test using the GF freeway data compared with the well known california and filter algorithm, Wavelet Algoritm has the potential to achieve significant improvement in incident detection performance.
出处
《西安公路交通大学学报》
CSCD
北大核心
1997年第02b期134-138,共5页
Journal of Xi'an Highway University
