Design And Implementation Of Complete Coverage Path Robot Based On ROS

With the development of artificial intelligence technology in the world,intelligent robots are gradually becoming hot.In the research of related technologies of mobile robots,path planning has become the research frontier and hot spot in the field of mobile robots,and it is one of the keys to distinguishing mobile robots from fixed robots.It is also a key technology for its intelligent and autonomous mobility.Full coverage path planning is an important branch in path planning research.At present,the full coverage robot basically uses unit segmentation,and the cattle farming segmentation is a mature algorithm in the unit segmentation.The irregular obstacles can not be correctly divided into regions,and the repetition rate of the robot full coverage path is increased.In the development of the fullcoverage robot system,due to the division of labor in the robot industry,the various components that make up the robot come from different manufacturers,which brings trouble to individual development.In view of the above problems,this paper designs a low-repetition full-path coverage robot system based on ROS robot operating system.Aiming at the problem that the segmentation of complex environment is inaccurate in the full path coverage algorithm,an improved full path coverage algorithm is proposed and implemented.The slice segmentation strategy is adopted,and the laser radar sensor is widely used to measure the wide range.The entire area is covered by a polygonal line path,and the A* algorithm is used for path planning between the areas,so that the walking path is more optimized,and the minimum repeated path length and larger coverage area are obtained.A multi-sensor distributed control robot system is proposed.For a single sensor to measure the surrounding environment in a complex environment,the encoder and the laser radar are used to obtain more accurate depth information and add a collision sensor.In the event of an accidental collision of the robot,it can be handled accordingly.The main body adopts distributed control,which is easy to manage.The bottom module controller adopts the cost-effective stm32 processor,and the upper layer adopts the Raspberry Pi to carry the ROS operating system,and uses its loosely coupled mechanism to build the robot software framework to improve code reuse.Reduce the difficulty of development.In the aspect of robot construction,the hector slam algorithm based on Gauss-Newton method and the RBPF algorithm based on particle filter are compared and analyzed.The improved algorithm gmapping is used to establish the grid map,so that the robot can achieve full path coverage on the map.Finally,basic functions such as robot construction,dynamic and static obstacle avoidance,and path planning are verified in the simulation environment.Through the operation of the robot in the actual environment,the coverage of the full-path coverage robot is 88.96%,and the repetition rate is 14.87%.