Pose Detecting Of A Parallel Robot Based On RBF Neural Network

6-DOF parallel robot do not only big in rigidity, small in bulk, high in moving precision, big in ratio of force or moment of force, good in dynamic performance, easy for feedback control, but also can actualize the space movements in six degrees of freedom. So it becomes a very important tool which is widely used in many dynamic simulative experimental apparatus, including airplane, ships and warships, space navigation and automobile, and it is an important tool for the development of airplane, ships and warships, space navigation and automobile.Detection of the moving platform position and orientation is always one of the key issues which troubled substantive breakthrough in the study of. So in this paper 6-PTRT parallel robot is selected for its research on moving platform position and orientation detection.Firstly, the kinematics of the 6-PTRT parallel robot is analyzed, for solving the homogeneous transformation matrix to complete coordinate transformation and completes the 6-PTRT parallel robot Jacobian matrix. Secondly, according to parallel robot has the feature of easy inverse kinematics solution but hard forward kinematics solution,6-PTRT parallel robot inverse model is established, then completed the 6-PTRT parallel robot inverse kinematics solution. Through the successive approximation of mobile link moving distance, this paper gets a kind of numerical algorithm for normal solutions. For its complexity of numerical algorithm calculating, and relying on the initial value, and RBF neural network functional approximation has arbitrary precision capabilities, small size and fast convergence properties, this paper established a method of Newton iterative normal solution. RBF is used for nonlinear mapping, resolving the initial value Newton iterative, and it can effectively avoid the process of kinematics solution complex calculation with successive approximation. Thirdly, the kinematics programming of parallel robot is completed, so does the control system design of 6-PTRT parallel robot, in which the upper monitor is IPC computer, lower monitor is PMAC motion control card, and the AC servo motor is used for drive the system, to control the parallel robot in real time.Finally, experiment is completed for test the algorithm of position and orientation detecting. The experiment indicates that the 6-PTRT parallel robot can realize the detecting of kinematics position, and the accuracy requiring can be realized.