Trajectory planning based on variable structure GA of a three-limbed robot

A novel progressive genetic algorithm is developed for motion planning of a three-limbed robot. The proposed motion planning method can be used to find a optimal joints trajectory from the initial to the final position and orientation. On the basis of the genetic algorithm a kind of variable structure genetic algorithm is proposed to solve the problem of motion planning of the three-limbed in dynamic environments. The variable structure genetic algorithm changes the original structure by abandoning Elitist Model, expectation selection, reproducing population and changing the probability of crossover and mutation. Experiments results show that the former algorithm is effective in static environments and the latter algorithm is good at dynamic environments.

I. Introduction

An increasing interest in the development of special climbing robots has been witnessed in last decades. The motivations behind it are to increase operation efficiency and protect human health and safety in dangerous tasks, such as cleaning high-rise buildings, spray painting and sand blasting of gas tanks, inspecting and maintaining nuclear facilities. Climbing robots, with their capabilities to adhere to wall surfaces and move around carrying appropriate sensor or tools, are able to replace human workers in these dangerous duties and eliminate costly erection scaffolding [1].