Research On Indoor Map Construction And Navigation Technology Of Pharmaceutical Loading Mobile Robot

The pharmaceutical manufacturing industry is related to the stable development of national health and society.China has the world's largest pharmaceutical market and has a huge demand for pharmaceutical manufacturing equipment.In order to meet the aseptic pharmaceutical production,pharmaceutical loading robots play an important role in the production of pharmaceutical automation.Traditional pharmaceutical loading robots have low intelligence,orbital or patrol-type navigation methods and indoor positioning methods based on WIFI and UWB,all require manual layout of mobile scenes,which requires a lot of cost and system scene adaptability.low.In view of how intelligent mobile robots move autonomously in an indoor environment without a priori information,this paper will study and implement indoor positioning and mapping of pharmaceutical loading mobile robots and path planning algorithms based on existing maps.For indoor positioning and map construction,this paper analyzes,introduces and compares the advantages and disadvantages of Gmapping,Hector and Cartographer algorithms with the laser SLAM algorithm.Based on the Hector algorithm,it is based on the Hector algorithm.A laser SLAM method based on inertial navigation for angle compensation of lidar data is proposed.The method uses a Gauss-Newton-based scan matching method,so that the calculation of the robot position is no longer dependent on the odometer model.At the same time,the extended Kalman filter is used to obtain accurate angle information and compensate the laser data to improve the accuracy of constructing the map.The experimental results show that the proposed algorithm reduces the angular error of the map wall and the error of the robot position matching,which proves that the method has good scene adaptability and practicability.For the navigation algorithm of the mobile robot,according to the state of the obstacle in the environment,it can be divided into static(global)path planning based on existing maps and local(dynamic)path planning based on sensor information.This paper studies and implements three global path planning algorithms,Dijkstra,A* and RRT,and two local path planning algorithms,artificial potential field method and dynamic window(DWA)algorithm.The experimental analysis shows that the A* algorithm has shorter path distance and reasonable computational cost,and the dynamic window method is not easy to fall into local optimum,which can well meet the indoor working scene of the pharmaceutical loading mobile robot.Aiming at the problem that the A* algorithm planning path is too close to the obstacle,an improved A* algorithm based on the potential field function is proposed.By adding the potential field function to the heuristic function to detect the obstacle distribution within the safe distance of the robot,the obstacle is made.The surrounding F value increases,and the search algorithm avoids the point where the F value is large to plan a safe path.In addition,the search factor reduction algorithm is introduced to take time,the direction vector between the robot and the target point and the potential field function vector are subjected to point multiplication,and the negative result is summed,and the distribution of obstacles in the current search direction is predicted by the search factor.,saving algorithm search time.The experimental results show that the proposed improved method can effectively improve the security and real-time performance of mobile robot navigation process.Finally,the improved laser SLAM algorithm is used to construct the indoor positioning map.Based on the existing map,the AMCL algorithm is used to locate the mobile robot.The improved A* algorithm and dynamic window method are used to complete the navigation algorithm of the mobile robot.Experiments show that the system can realize that the mobile robot reaches the target point without collision in the indoor environment.