基于强跟踪滤波的全自主机器人目标预测

全自主机器人在目标跟踪过程中,为了克服视觉系统造成的决策延时,需要对目标进行预测。目前全自主机器人目标预测方法对目标模型有强依赖性,并对突变状态的预测滞后。本文提出将强跟踪滤波理论应用于全自主机器人目标预测,通过引入渐消因子,克服了其它目标预测方法的缺点。仿真结果说明强跟踪滤波对目标的预测比较精确,并对突变状态反应灵敏,说明该方法的有效性。

Object Prediction of Autonomous Robot Based on Strong Tracking Filtering

In the object tracking of autonomous robots,it is necessary to predict the object position to overcome the decision-making delay,which is caused by the vision system.But there exists strong dependence on the object model and delay of the mutational states in object prediction methods used presently.In this paper,the theory of Strong tracking filtering(STF) is applied in the object prediction of autonomous robots to avoid the disadvantages of other methods by introducing fading factors.Simulation results show that the object prediction using STF is relatively precise and is sensitive to mutational states,thus proves the validity.