Path planning of a snake-like robot based on serpenoid curve and genetic algorithms

Abstract

The path of the snake-like robot has no repetition because its motion is influenced by manifold indeterminate factors such as ground condition, mechanism's vibration and motor voltage's variety. A novel path planning technique based on serpenoid curve and genetic algorithms (GAs) has been proposed to control the snake-like robot in Shenyang Institute of Automation (SIA, China). First, the ranges of the path and the curvature deviation have been calculated approximately and set as the bounds of genetic algorithms. Then using real time dual genetic algorithms, this path planning technique not only can decide the shortest path and the minimum curvature deviation, but also can limit the motion error's influence. Simulation results show that the results of the second layer of GAs are more available than that of the first layer of GAs and this novel technique is effective for the path planning of the SIA snake-like robot.