改进IMM算法在机器人目标跟踪中的应用

摘要

在机器人对运动目标跟踪的过程中，由于目标运动状态具有多样化的特点，不能运用单一的运动模型对其进行跟踪。文中将当前统计模型(CS)和匀速模型(CV)交互，并自适应调节“当前”统计模型中的目标加速度和模型之间的转移概率，形成新的CVCSIMM算法，使其能够更有效地反映目标的机动特性。在Matlab上对本算法进行了仿真研究，并与基本CVCAIMM算法、改进前的CVCSIMM算法进行了比较。 Monte Carlo仿真结果表明：本算法减小了跟踪过程中的误差，提高了对机动目标的跟踪精度。%In the tracking process of robot for moving target,since the target motion state has diversified characteristics,can not use a sin-gle movement model to carry on the track. In this paper,the current statistical model (CS) and uniform model (CV) is interacted,and the target's acceleration of the"current" statistical model and the transition probability between the models are adjusted adaptively to form a new CVCSIMM algorithm, so that it fully reflects the maneuvering characteristics of the target. The Monte Carlo simulation is carried out using Matlab software,which is compared with basic CVCAIMM algorithm and the old CVCSIMM algorithm. The results show that the proposed algorithm reduces the error in the tracking process and improves the tracking precision of maneuvering target.