Design And Implementation Of Path Planning Algorithms And Experimental Platform For Serving Robots

As the mature of the market of serving robots,users demand more for robots which generates more cases for multi robots working together.Tradi-tional slowly path finding algorithms could not satisfy the need of users,which motivates the research on sub-optimal path finding.Meanwhile,when it comes to multi agent path finding problems,robots could easily conflict with each other,form a deadlock or incur the memory explosion.All these problems should be solved through the improvements of traditional algorithms.How-ever,the current fashionable robot platforms are a bit complex.Based on these,this paper mainly researches the following three parts.Firstly,to improve the single robot path finding algorithm,we abstract the current newly conditions for sub-optimal path finding and research on a method of extending in poly-line,which improves the traditional sub-optimal path finding algorithm.Secondly,to improve the multi-agent path finding algorithm,we research on the flow annotated method,and abstract the general rules for flow annotated,proposing a method to verify the connection of map.On the basis of these,we propose a solution for multi-agent.path finding.Thirdly,this paper design and conduct a platform for path finding simu-lation.To better experiment the traditional path finding algorithms especially the flow annotated algorithms,this paper draw the front end of platform and build the framework for path finding algorithms.Through the Test-Driven-Developing concept,this paper finished the test and framework of platform.