Robot path planning using improved genetic algorithms
Shao Guo · Jisuanji gongcheng yu sheji · 2008
The decomposed robot's three-Dimensional workspace is discreted and modeled by grids.A kind of random fast searching method,based on the rule of the grids' array number,is designed to create the initial population.The genetic algorithms operators is improved properly to avoid degeneration,which will make the algorithm converge faster.Results of the simulation which uses path length and energy costs as the criterion show that the improved algorithm is simple but effective,and also has a strong ability to adapt to the changes of the robot's workspace.