Application of Singular Value Decomposition Methods For Optimization of Fixed Parameters Involving Resolved Rotational Motion Rate Control In A Motion Simulator

Abstract

The application of Singular Value Decomposition (SVD) methods to investigate motion control of a robotic simulator adds new insight into the optimization of fixed parameters within the device. These fixed parameters within the simulator can be optimized apriori before motion field simulation actually begins by the investigation of the SVD of the Jacobian matrix. Of particular interest is the minimum singular value which represents the ratio of end effector velocity to joint velocity in the direction most difficult to move. Using a Damped Least Squares approach, the sensitivity of the fixed parameters is studied with no weighting on the joint velocities (λ = 0) compared to the situation when these joint velocities have a damping coefficient (λ(t)≫ 0).