Development of a general manipulator path planner using fuzzy reasoning

Abstract

Purpose Aim for efficient motion planning of industrial robot with high degree of freedoms in both static and dynamic environments. Design/methodology/approach A multi‐agent based general path planner for serial manipulator is proposed in this work. A hierarchical structure developed based on fuzzy reasoning is employed in the planner. The high level in the hierarchical structure is designed to dynamically assign each link an appropriate behaviour and the low level is designed to determine the joint speed according to the behaviours assigned by the high level. Findings Combination of multi‐agent concept and fuzzy reasoning approach can obtain both flexibility and efficiency in motion planning of serial manipulator with high degree of freedoms. Research limitations/implications Multiple local minima problem occurred in complex manipulation scenario has not yet been considered. Practical implications Applicable for real time motion planning of serial industrial robots with high degrees of freedom in 3D space. Originality/value In this research work, we make use of the of multi‐agent plus fuzzy logic concept to design a novel manipulator motion path planner, in particular, we introduce a novel mechanism into the fuzzy logic algorithm with the “back‐tracking” ability to avoid the local minima problem.The proposed motion planner has advantages on low computational cost, the suitability for real time path planning in 3D space and the capability to escape simple local minima.