Genetic algorithm for dynamic path planning

Abstract

Optimization in dynamically changing environments is a hard problem. Path planning for mobile robots is a complex problem that not only guarantees a collision-free with minimum traveling distance but also requires smoothness and clearances. This paper presents a genetic algorithm approach for solving the path planning problem in stochastic mobile robot environments. The genetic algorithm planner (GAP) is based on a variable length representation, where different evolutionary operators are applied. A generic fitness function is used to combine all the objectives of the problem. In order to make the algorithm suitable for both static and dynamic environments, problem specific domain knowledge is used.