A path planning approach to moving robot based on genetic algorithms

Abstract

A path planning approach to moving robot in dynamic environment based on genetic algorithms is presented with the consideration of the shape of robot and obstacles in the case, when the obstacles and the target are moving in a certain field. For the path planning, genetic algorithms are used in twice. Two simplified and perfect fitness functions are derived for twice genetic algorithms. The results of simulation experiment show that the method can well enhance a moving robot navigation to its moving target when it faces new scenario on-line. This method can also be used for automatic navigation of intelligence robot in the case that the target is moving.