Research On Control System And Path Planning Of Move-in-mud Robot

Move-in-mud robot is a new-type and special-use underwater robot. Now, the main method to salvage the wrecked vessel is to using buoy. Punching a hole in the mud under the sea is the key technology of salvaging the wrecked vessel using buoys. At present, divers finish the work of punching a hole manually. The work is laborious, inefficient and dangerous. The move-in-mud robot can perform the hole excavating work to benefit sling passing in the planned trajectory in the mud under water. The control system and path planning are the essential technology in the move-in-mud robot' s research, which is the premise and ensurence for the robot to accomplish the hole excavating work through planned trajectory. With the support of "Basic Technology Research on move-in-mud Robot Based on Worm Principle" by the National Natural Science Funds, research on move-in-mud robot' s control system and path planning are developed in this paper.The development and present situation of the mobile robot, its method of path planning are summarized in the paper. The development of the control theory, the method of intelligent control and present study situation in mobile robot domain are also introduced in the paper.Kinematics and dynamics of the move-in-mud robot have been studied. Aimed at the vermiculation characteristics, kinematics model of the move-in-mud robot is established. Then the forces on the move-in-mud robot under the ocean' s mud are analyzed. At last, the dynamics equations of the move-in-mud robot are established.Because of the strong uncertainty of the move-in-mud robot' s working environment, a kind of adaptive PID control method using BP neural network is proposed. The BP neural network is used to modify the parameter and makes the move-in-mud robot have adaptive capacity.The neural networks model of the head movements of the move-in-mud robot is established by investigating the head movement model of the move-in-mud robot and it is designed to provide grads data for the BP neural network learning.The path planning method for the move-in-mud robot is investigated in this paper. A kind of local path planner based on fuzzy neural network is proposed in this paper. The path planning strategy is that: firstly, layout the off-line global path; then, when the robot is moving, the local path planner gets the sensors' data, makes on-line planning, modifies the real-time path and decides the sub-goal.The control system of the move-in-mud robot prototype is designed in this paper. Based on the mechanical characteristic and special working environment of the move-in-mud robot, two level computer control scheme is adopted. The PC computer is chosen as the host, and the AT90S4433 single-chip is chosen as the slave. The serial communication is used for exchanging data between the host and the slave. With the experiment platform established in laboratory, the experiment researches on the move-in-mud robot prototype are made. The experiment results show that the control scheme proposed is feasible. The prototype makes stability foundation for engineering sample of next step.