Optimal Motion-Cueing Algorithm Using Motion System Kinematics

Current motion-drive algorithms tend to minimize a cost function to provide best sensation while maintaining the motion system within its constraints. The cost functions comprise terms that express motion system constraints in operational space. In the present paper, a novel approach is taken which directly applies the physical constraints to where they basically exist, i.e., to joint space. Actuator lengths and their rates are expressed in terms of system states and then considered in the cost function, then formulated conveniently so as to be addressed by optimal algorithm. Not only does the method provide rational values for initial choice of weight matrices but also it clearly increases the fidelity of motion system response by effective usage of workspace. It assures maintaining actuator lengths and rates within their limits while providing required motion sensations. Further, it degrades the need for scaling and final soft-limiting filters and also eases the filter tuning process which requires the experience of a motion cueing expert or conducting large series of experiments to be carried out. In addition, the overall effect of simulator motion on motion system constraints is considered in a coupled manner while being of significantly reduced computational load. This modification also leads to better performance of simulator in constraining actuator extensions in coupled maneuvers and consequently better assurance of the safe operation of simulator.