抗野值卡尔曼滤波器在排队长度预测中的应用

Application of anti-outlier Kalman filter in prediction of queue length

车辆排队长度是智能交通控制系统中最关键的交通参数之一,如何准确、有效地预测车辆的排队长度是近年来排队长度预测中研究的主要问题。为此,在利用卡尔曼滤波理论建立排队长度预测模型的基础上,针对滤波过程中较容易出现的野值问题,分析了野值对标准卡尔曼滤波的影响机理,采用了基于M估计的卡尔曼滤波野值处理方法。仿真结果表明,所提方法可有效消除野值对滤波的不利影响,提高排队长度预测的精度,具有较好的应用价值。

The queue length is one of the important parameters in intelligent transportation control system. So, how predict the length of the queue of vehicles accurately and effectively is the main issue of he prediction of the queue length in rencently years. On the basis of setting up the queue length prediction model with the Kalman filter theory is presented. For the common problem of outlier in the process of filtering, the influencing mechanism of the outlier on the standard Kalman filter is analyzed and the solution is proposed to dissolve the outlier with the Kalman filter based on the M-Estimation. The simulation results show that the solution proposed can effectively eliminate the negative effect of the outlier on the filter and can consequently increase the precision of the prediction of the queue length.