公交调度系统事件检测算法研究

Research of Accident Detection Method of Bus Dispatching System

利用小波Mallat算法对GPS采集的速度信号进行分解,并利用小波变换模极大值与信号奇异点的对应关系对信号进行突变点检测.根据小波可同时进行时频分析的特性,将GPS采集的速度信号在时频空间展开,并根据不同的公交车运行状态分别对信号进行分析.最后提出了以突变点大小、位置、个数以及平均速度和速度零点时间长短为参数的综合事件判别条件.通过与实测公交运行状态对比发现,以小波变换对GPS采集的速度信号进行突变点检测是非常有效的.

Velocity signals collected by GPS are decomposed by Mallat algorithm of Wavelet. According to the relationship between the maximum wavelet coefficient modules and the singularity signal, the breaking dots are detected. Based on Wavelet＇ s Time-frequency analysis characteristics, the velocity signals collected by GPS are spread to time-frequency space, and different signals of different movement states are correspondingly analyzed. Considering the parameters of breaking dot＇ s size, position, number, mean velocity and zero velocity continuance time, the differentiate condition are finally proposed. Compared with the actual bus movement state, it is very effective to use Wavelet to analyze the signals collected by GPS.