Research On The Localization And Full Coverage Path Planning For A Cleaning Robot

As one of the most typical representatives of service robots,cleaning robots can accomplish cleaning tasks independently and intelligently,which has a very broad market prospect.The key technologies such as obstacle avoidance,positioning,map building and path planning,are the hotspots of robot research.Among them,positioning and full coverage path planning are the most representative technology.Aiming at the problems of low positioning accuracy and low coverage and high repetition rate,this paper studies the positioning method and full coverage path planning technology of cleaning robots.Firstly,the overall scheme of the cleaning robot is designed.Through the analysis and comparison of the current common methods,The positioning and path planning scheme of this subject is determined.In addition,the hardware system design of the cleaning robot has been completed,including the design of control system,driving mechanism,cleaning mechanism,sensing system and power supply system.Secondly,based on the kinematics model of dead reckoning,the positioning algorithm based on encoder and gyroscope is realized.Experiments show that there are some problems in location based on encoder alone,the positioning accuracy of the robot in turning process can be improved by adding gyroscope,The data transmission between the robot and the host computer is realized based on WiFi module,the function and interface of the host computer are designed by using MATLAB,thus the effective control of the robot and the real-time display of the location trajectory can be realized.Then,in order to solve the problem of low coverage and high repetition rate of the cleaning robot,the element decomposition method is used to achieve full coverage of the environmental area.Based on the known map information and the idea of rectangular decomposition,the map unit is decomposed by grid method.The connection order of each sub-region is determined by depth-first traversal of the map.A* algorithm is introduced to plan the shortest connection path of adjacent sub-regions,and the A* algorithm is simulated.Circuit traversal is adopted as the basic traversal method,and the situation of boundary obstacles and isolated obstacles in the environment is analyzed.In order to realize the path planning between two points,a path planning method based on PID control is proposed.The distribution of target points in the environment is studied,and different path planning strategies are designed.Finally,based on the built experimental platform of cleaning robot,various experiments are carried out to verify the effectiveness of the algorithm.Including the test experiments of each module,the local area traversal experiments of boundary obstacles and isolated obstacles in the environment,straight-line walking experiment and straight-line turning experiment.The experimental results show that the circuitous traversal strategy adopted in this paper can improve the coverage rate,and the path planning algorithm between two points based on PID control can effectively improve the accuracy of fixed-point planning.