Design of adaptive moving-target tracking control for vision-based mobile robot

Abstract

This study constructs an adaptive moving-target tracking control (AMTC) scheme via a dynamic Petri recurrent-fuzzy-neural-network (DPRFNN) for a vision-based mobile robot with a tilt camera. First, a continuously adaptive mean shift (CAMS) algorithm is adopted for the moving-object detection, and a model-based conventional sliding-mode control (CSMC) strategy is introduced. Moreover, it further designs a model-free AMTC scheme with a DPRFNN for imitating the CSMC strategy for relaxing the control design dependent on detailed system information and alleviating chattering phenomena caused by the inappropriate selection of uncertainty bounds. In addition, a switching path-planning scheme plus the AMTC is designed without detailed environmental information, large memory size and heavy computation burden for the obstacle avoidance of a mobile robot. Furthermore, numerical simulations are given to verify the effectiveness of the proposed AMTC scheme under different target tracking, and its superiority is indiented in comparison with the CSMC System