Research On Control Strategy Of Sea-Borne Self-Stabilizing 6-Dof Parallel Mechanism

Under the circumstances of exploitation of marine resources,rescues and military operations,it's usually required that the devices involved be isolated from the disturbances generated by sea waves in order to keep the devices still in certain degrees of freedom relative to the inertial frame.For light loads,electric serial self-stabilizing mechanisms are often enough to keep those loads stable with satisfying accuracy.While for heavy loads,electro-hydraulic servo parallel self-stabilizing mechanisms are widely researched due to advantages like fast responses,great bearing capacity and accuracy in motions.This dissertation focuses on kinematics and dynamics of standard Stewart parallel mechanism,measurement and estimation and simulation of disturbances induced by sea waves,prediction algorithm and simulation of disturbances,control strategy of the platform and finally the test for it.The dissertation first chose standard Stewart parallel mechanism as the ultimate structure of the self-stabilizing platform because of its bearing capacity;then the relationship between its workspace and the parameters was studied as well as its kinematics;after a set of parameters was chosen,dynamics and its model under Simulink was also analyzed.To keep the platform at a desired pose,estimating the disturbances of the ships induced by waves is necessary.An algorithm based on IMU measurements and quaternion theory was designed and simulated,an improvement of the algorithm was also mentioned;for the simple translation disturbances,a filter specially designed for filtering accelerometer measurements was too simulated and solutions for complicated situations were introduced theoretically.Then considering the lag induced by filtering and calculation during the process of pose estimation which significantly degrades the performance of the platform,AR model was researched and proved to be effective when it's used to predict wave motions;based on AR model,a real-time prediction algorithm was designed and simulated.Finally the control strategy of an electric-hydraulic servo actuator was utilized,and the control strategy of the entire system including pose estimation,prediction and the platform itself was designed and the control strategy was proved to be effective under Simulink.A two-story Stewart parallel mechanism was built to partially testify the control strategy.