Study On Localization And Path Planning For A Cleaning Robot

Localization and path planning, which are the basis of robotic intelligence and aotumotion, are two key aspects in the researches of cleaning robot. This paper deeply analyzes and discusses the localization and full coverage path planning method of cleaning robot.First, principles of the four sensors adopted in this paper are introduced. The methods that indoor localization is realized by fusion of ZigBee, laser sensor and encoder, and outdoor localization is conducted by fusion of DGPS, laser sensor and encoder are proposed to realize the accurate localization both indoor and outdoor.Second, based on laser sensor a simple method used to detect the road edge is proposed, which can realized the detection and lineshape fitting of the road edge. Also, the satic and dynamic obstacle avoidance algorithm is introduced.Third, since the full coverage path planning study is mature in the indoor structuring environement, and few researches in outdoor situation, an improved method based on biologically inspired neural network is proposed to realize outdoor path planning. The computer simulation shows the menthod feasible.Finally, according to the requirement of the cleaning robot localization and full coverage path planning, a set of control system is designed including the subroutines of each control module and the man-machine interface. The experiments are conducted by the cleaning robot prototype, and the results show the methods proposed in this paper to be accurate and effective.