Improved Genetic Algorithms Based Path planning of Mobile Robot Under Dynamic Unknown Environment

Abstract

Aiming at path planning of mobile robot under part of dynamic unknown environment, there are some shortages in the aspects of produce of initial population and the structure of specific genetic operator in current used genetic algorithms. In this paper, using the position feedback and forecast of moving direction of obstacle, we present a new method of robot path planning based on improved genetic algorithms combined with numerical potential field. The problem of path planning and avoiding obstacles under dynamic environment was resolved by path re-planning. The shape of obstacle is not limited, and the research is close to the real work environment of robot. The specific genetic operator, fitness function and real coded were designed in this paper. The simulation instances under multi various complex dynamic environments verify that our algorithm of robot path planning is high efficient, and the operation speed and accuracy are improved. http://www.ieee-icma.org