Study On Inverse Kinematics And Path Planning In Ultrasonic Testing For Curved Surface Parts

Ultrasonic non-destructive testing methods are widely sued for detecting flaws in materials. It's a frontal research to realize automated test for the complex curved surface parts. Combined with the techniques requirement in the automated ultrasonic testing system, some key techniques in relation to inverse kinematics and path planning of the ultrasonic test robot are studied in this dissertation.In chapter 1, the background of the research work and significance of this dissertation were discussed. Some crucial techniques involved with the ultrasonic testing for curved surface parts, inverse kinematics and path planning of ultrasonic test robot, were analyzed. Finally detailed research contents and given chapters arrangement.In chapter 2, some techniques involved with for the path planning of ultrasonic testing robot were analyzed, such as the grid modeling, collision detection method based on hierarchical bounding box and the A* heuristic search Algorithm.In chapter 3, the inverse kinematics of the ultrasonic testing robot with redundant degrees of freedom was studied. Kinematics modeling of the ultrasonic testing robot is the basis of automated ultrasonic measurement and testing. The robot's mechanical structure was analyzed, and the robot's inverse kinematics expressions were derived. Then, two inverse kinematics analysis algorithms were proposed based on analytical solution combined with fixed joint angles. The advantages, disadvantages of these two algorithms and their applications were discussed. Finally, several experiments showed the validity of the methods of from surveying accuracy and computing time.In chapter 4, the path planning of the ultrasonic testing robot was studied. To avoid collision with the curved surface part during ultrasonic probe's long-distance movements, a method of collision-free shortest path planning was proposed. The robot's working environment model was established based on the grid modeling and the hierarchical bounding box for collision detection. Then, the collision-free shortest path from the starting point to the target point were search out, using the A* algorithm. Finally, this path planning method was tested with a series of examples, and some key parameters of this method were compared.In chapter 5, the dissertation was summarized, and the possible research in future is presented.