A non-geometric method for inverse kinematics solution of a delta-structure ship motion simulator and its trajectory control

Abstract

This paper presents a non-geometric method for solving the inverse kinematics of a ship motion simulation system having Delta structure as shown in Fig. 1, and then investigated the problem of path tracking control generally. As the complexity of the joint structure of this system, obtaining the inverse kinematics is difficult and computationally expensive and traditional methods such as geometric, iterative and algebraic seem not completely effective. So, author uses the ability of ANFIS (Adaptive Neuro-Fuzzy Inference System) to learn from training data, it is possible to create ANFIS with limited mathematical representation of the system. Simulation is carried out by Matlab to indicate that this method has the advantage of faster learning rate, higher identifying precision and better real-time ability. The purpose of trajectory control is to define a temporal motion law along a given geometric path. A method for trajectory control is presented which is based on a modified trapezoidal velocity profile (TVP), and velocity control method is proposed using a PID controller to ensure the end effector tracking the desired path with high precision. At last, a demo model is used to verify the feasibility of the method. The results of the paper are very useful for solving the inverse kinematics of manipulator as well as demonstrate the improved performance of controller combining between TVP and PID over conventional trajectory control - point to point type.