Researches On The Global Path Planning Of Mobile Robots

Path-planning is a very important task in robot navigation system. Global path-planning problem is an important issue on research of mobile robots. And the research of Global path-planning is very meaningful.How to find a non-obstacle path from start to target is the focus of this paper. The author summarizes some methods of path-planning and refers to the latest fruit on path-planning technique. The following is the author' s works :Firstly, the basic idea and detailed description of implementing robot path-planning is analysed. On the basis, this paper presents a fast neural network algorithm. The path-optimization function and simulated annealing function are designed, then a simulation and a convergence analysis are performed.Secondly, the basic idea of implementing robot path-planning with genetic algorithm is discussed. The new concept of interruptive non-obstacle is introduced. Adopting real encoding and genetic simulated annealing algorithm , the coding scheme, selecting strategy .genetic operators and stopping rules are designed. Then a simulation and algorithm performance testing are performed. This paper improves the genetic algorithm, so the shortage of path-planning with genetic algorithm is overcomed.At last, this paper primarily discusses the robots path-planning in uncertain information. Some correlative problem , such as navigation construction, map-building, uncertain information discription, is discussed.