Mobile Robots Path Planning Optimization in Static and Dynamic Environments
By Ahmed Shamli, August 2004
Abstract:
Path planning for mobile robots is a complex problem. The solution should not only
guarantee a collision-free path with minimum traveling distance, but also provide
a smooth and clear path. In this research, a Genetic Algorithm Planner (GAP) is
proposed for solving the path planning problem in static and dynamic mobile robot
environments. The GAP is based on a variable-length representation. A generic fitness
function is used to combine the objectives of the problem. Different evolutionary
operators are applied some are random-based, and others use problem-specific domain
knowledge. Various techniques are investigated to ensure that the GAP is appropriate
for dynamic environments. To further increase the efficiency of the GAP, an
Island-based GA (IGA) is developed on a ring topology and Message Passing Interface
(MPI) library is utilized to implement the IGA. A new Local Search (LS) is also
developed in this research and different approaches are examined for combining the
LS algorithm with the GAP to obtain superior solutions.
