Multi-objective Genetic Optimization And Its Applications, Robot Path Planning

Path planning is the key technology of Mobile Robot research., the research of path planning is very important, especially when the hardware of robots cannot reach a high precision in short time. This paper does research in Multi-Objective Genetic Algorithm and its application in robot path planning in static mobile robot environments.The basic theory of GA and the research status of path planning are summariaed in this paper, and a new algorithm is expatiated based on Clustering method and improved elitist strategy. To solve the dissatisfied representation of NSGAⅡwhen it is used in high dimensionality, the new algorithm uses Clustering Method instead of crowding procedure, which is used in NSGAⅡ; then brings the non-elitist individuals into crossover operation to overcome the limitation of elitist strategy, which is aim to assure the constringency and improve the diversity of algorithm. Then developed the model of mobile robot environment, and the new algorithm is applied to the robot path planning, some key technologies are used in creating initial generation and parameters setting in order to get better experiment result.The simulation results indicate that, compared to NSGAⅡ, in the same environment, the new algorithm can search out more non-dominated individuals which are the feasible paths, and improves the diversity of the algorithm.