Wheeled Robot Path Planning in Natural Environment

In this work we present a method to obtain the obstacle-free path optimized under shortest travel time condition in a a priori known map, with velocity profile generation. The method is in fact a set of three submethods: the pathfinding algorithm (PA), the optimization algorithm (OA) and the velocity profile generation algorithm (VPGA). The PA is used as a subfunction for the VPGA genetic algorithm that searches for the PA parameters which give the time-optimal path. As the last step the VPGA algorithm is used to obtain the velocity profile for the robot. The velocity profile takes into consideration areas where the path is close to the obstacle. In those areas the robot velocity is reduced to make it less sensitive for errors that can cause a collision. The velocity profile is not computed in the GA fitness function due to computation cost. The solution fitness is approximated using extracted path characteristics, which approach is much more effective.