Research On Movement Law And Control Methord Of 6-DOF Parallel Platform

The 6-DOF parallel platform has many advantages of large bearing capacity, good rigidity, high precision, fast system dynamic response, without accumulative error, etc. So it is widely used in many fields, such as robot, flight simulator, new type machine, air space docking, and in this paper the 6-DOF parallel platform is used as an equipment to simulate the ship moving in the sea. Compared with the actual sea tests, it has advantages in being controllable, non-destructive, economical and reliable. The 6-DOF parallel platform can provide six types of motion as heave, portrait motion, transverse motion, pitch, roll and yaw. Through the simulation of various sea conditions, it can guide to research the kinematics and dynamics of the ship.It applies the mathematic model of the 6-DOF parallel platform with matrix analysis and homogeneous transformation, and analyzes the velocity and the acceleration of the platform; the dynamics model of the platform is based on driving to the bar for additional force and additional torque, d'Alembert principle and the principle of virtual work. It simulates the kinematics and gets the curve of movement of all hydraulic cylinder, and researches the characteristics of different motion.It analyzes the inversed kinematics of the 6-DOF parallel platform, and sets up the model of neural network to convert the vector of the platform length into the vector of the position for non-uniqueness of analytical solutions of forward kinematics. It gets input and output data as training sample from inversed kinematics, realizes the nonlinear mapping from the joint variable space to the operation variable space for the platform, so the position of the 6-DOF parallel platform can be obtained by the importation.Based on the analysis of the traditional PID control, it proposes PID control based on the neural network and simulates this control strategy. It can learn and tune the PID parameters by itself with the neural network online. The result shows that this control method can achieve the tracking of the system, and has good dynamic and static properties.