Research On Navigation And Grab Algorithm Of Mobile Robot

As a key carrier of intelligentization in the new century,mobile robots have gradually been widely used in industries,services,agriculture and aerospace.However,with the increase of task requirements and environmental diversity,the research of mobile robots is constantly facing new challenges,especially its autonomy and informatization.Based on MT-R mobile robot platform,this paper studies the navigation and grabbing problems of mobile robots.The details are as follows:(1)Based on monocular vision,a target recognition and ranging algorithm is studied.First,according to the small hole imaging model,the camera is calibrated by Zhang's calibration method to obtain the internal parameters of the camera.Second,by using hough gradient detection algorithm and the edge detection algorithm,a specific target is found.Finally,after using the principle of geometric similarity and introducing the area characteristics,the relationship between the target distance and the image size is derived.The experimental results show that the algorithm error is reasonable and meets the reliability and real-time requirements.(2)Studying the reactive obstacle avoidance algorithm of mobile robot during navigation.First,a kinematic model of the wheeled robot is established and then the relationship between the wheel speed difference and the turning radius is derived.Based on the kinematics model of MT-R robot,a reactive obstacle avoidance algorithm for mobile robot based on fuzzy neural network and arc trajectory is proposed in this paper.Finally,after designing network structure and establishing CAD samples,the proposed algorithm is simulated on the ROS platform by using STDR mobile robot simulation software.What's more,physical verification is also performed.The experimental results show that the controller is quite effective when it navigate the mobile robot in an unknown environment.(3)Proposing an improved multi-objective particle swarm optimization algorithm for the six-degree-of-freedom simulation manipulator system.First,establishing the kinematical model for a 6-DOF simulation manipulator in joint space.Second,designing a trajectory plan that satisfies the kinematic constraints.Then,after analyzing the optimization function and establishing a multi-objective optimization mathematical model of the manipulator,an improved multi-objective optimization algorithm that suitable for the manipulator is proposed.The algorithm updates the particles through two iterations to ensure the diversity and convergence of the solutions.Finally,the effectiveness of the proposed algorithm is verified in the simulation.