Open Collections

Abstract

This thesis describes the modeling, simulation, and control of an inverted, ceiling-mounted Stewart platform, which is designed to be a motion simulator. This hydraulically actuated Stewart platform is capable of providing 10 m/s2, 400 degree/s2 accelerations and 1 m/s, 30 degree/s speeds to a 250 kg payload. The issues of modeling and control of such a platform are addressed here. The inverse kinematics and forward kinematics are studied first. The platform rigid-body dynamics are derived based on the virtual work principle and then combined with the actuator dynamics to simulate the response of the Stewart platform given a pre-planned motion path. Design and implementation of the link-space controller are discussed and also validated using experimental data. Cartesian-space controllers are also addressed. Motion drive algorithms are finally addressed to complete the system's function as a motion simulator. When the controller is well tuned, the bandwidth of the system can reach about 9Hz along the vertical axis for a payload of about 140 kg.