摘要
针对基于Kalman滤波的跟踪方法需要对噪声特性和车辆的运动规律进行假设的不足,提出一种基于灰色预测模型GM(1,1)的运动车辆跟踪方法.该方法通过不断更新的灰色预测模型GM(1,1),挖掘出车辆的当前运动规律,从而对车辆的运动位置进行快速准确的预测;然后根据预测结果搜索出运动车辆,实现运动车辆的跟踪.试验结果表明,该方法在不需要假设的条件下,能够较快较好地实现车辆跟踪.
A vehicle traking methods based on GM（1,1） is proposed to eliminate the noise assumption and motion assumption which are demanded in tracking methods based on Kalman filter. The model GM （1,1） is updated and the position of the vehicle is forecasted accurately and quickly according to the updated model in each step. Then the vehicle can be found near the forecast position and the tracking can be realized. The experiment results show that the tracking is finished effectively and quickly without any assumption.
出处
《控制与决策》
EI
CSCD
北大核心
2006年第3期300-304,共5页
Control and Decision
基金
国家自然科学基金重点项目(60134010)
