Research On Dynamic Path Planning Of Desktop Manipulator In Chemical Analysis

In recent years,artificial intelligence has developed rapidly,and robots have become more widely used in the fields of production and life,especially in dangerous environments that are unfavorable to human safety.In chemical analysis experiments,toxic substances and harmful gases produced by chemical reactions will cause irreversible damage to the health of analysts.Therefore,the application of the robotic arm in the field of chemical analysis is of great significance.The path planning algorithm affects the efficiency and safety of the motion of the manipulator,and is one of the key technologies to control the motion of the manipulator.In the chemical experimental environment,the operating environment is relatively complex,and the use of traditional path planning algorithms to control the manipulator will lead to problems such as low work efficiency and difficulty in path search.Moreover,in the face of dynamic obstacles,the robotic arm cannot make a timely obstacle avoidance response,which will lead to the failure of the analysis experiment and even cause unnecessary losses.Aiming at the problems encountered in chemical analysis experiments,this subject has carried out the following research on the dynamic path planning algorithm based on the six-degree-offreedom manipulator arm:First,kinematic analysis of the robotic arm.The D-H method is used to establish the kinematic model,and the kinematic constraints of the manipulator are clarified;and the numerical method is used to solve the inverse kinematics of the manipulator.Finally,write the URDF file according to the manipulator parameters and apply it to Move It!The simulation is carried out to provide the basis for the research on the dynamic path planning algorithm of the manipulator.Second,build a dynamic obstacle perception system.The depth image is obtained through the depth camera,and the depth image is converted into the point cloud image in the base coordinate system of the manipulator using the in-camera participating transformation matrix.The point cloud image is converted into an octree occupancy map using the octree structure,and the experimental scene is acquired and updated in real time.Third,improved static path planning algorithm.An improved RRT* algorithm based on goal-biased strategy and heuristic optimization strategy is proposed.The algorithm is divided into a search phase and an optimization phase.In the search phase,the algorithm uses the target traction area sampling and the sunflower area sampling strategy to speed up the acquisition of the initial path;in the optimization phase,the initial path is directly optimized by the heuristic subset sampling strategy to reduce redundant computation and improve the efficiency of the algorithm.When using this algorithm for path planning,a less expensive collision-free path can be obtained in a relatively short time.Fourth,the discrete collision detection algorithm is improved,combined with the GJK fast calculation distance algorithm to realize the dynamic collision detection of the manipulator.While reducing the number of collision detections,it avoids the "tunnel effect" during the movement of the robotic arm,improves the dynamic collision detection efficiency of the robotic arm.Experiments show that the algorithm can quickly and accurately detect dynamic obstacles and perform obstacle avoidance operations in time.