Motion simulation platforms are mechanical devices designed to replicate the dynamics of a given vehicle. These devices are very attractive for training individuals as drivers, pilots or passengers. In the case of river boats, the simulator consists of a section of the boat (hull) mounted over a 3 DOF parallel robot with a passive mass compensator (3UPS + PU). If users have mobility in the hull, an uncertainty in the position of the upper platform’s center of mass is produced. This variation may generate excessive loads on the robot that can be prevented by an adequate placement of the hull over the robot. Dynamic calculations, based on measurements of the real boat in motion, are computed by numerical simulations in SimMechanics. Three methodologies are presented for optimizing the configuration of a boat simulation platform. First, a manual procedure is developed in which critical cases are intuitively detected and evaluated. Then, two multi-variable optimization algorithms are used to systematically obtain the best position and orientation (pose) of the boat section: Genetic Algorithms and low discrepancy sequences. The pose is the design variable; the average forces are the objective functions and the maximum difference between the average forces is the fitness function. The article describes the design problem, the proposed optimization methodologies and simulation results for the optimal configuration.
Deterministic methods for stochastic computing using low-discrepancy sequences
