Multi-sensor control for 6-axis active vibration isolation.

摘要

The goal of this research is to look at the two different parts of the challenge of active vibration isolation. First is the hardware that will be used to accomplish the task and improve performance. The cubic hexapod, or Stewart platform, has become a popular solution to the problem because of its ability to provide 6-axis vibration isolation with a relatively simple configuration. A number of these hexapods have been constructed at different research facilities around the country to address different missions, each with their own approach. Hood Technology Corporation and the University of Washington took the lessons learned from these designs and developed a new hexapod that addresses the requirements of the Jet Propulsion Laboratory's planned space borne interferometry missions. This system has unique mechanical design details and is built with 4 sensors in each strut. This, along with a real time computer to implement controllers, allows for a great deal of flexibility in controller design and research into sensor selection. Other unique design features include a very soft axial stiffness, a custom designed voice coil actuator with a large displacement capability and elastomeric flexures both for guiding the actuator and providing pivot points on each strut.; The second part, and the primary area of this research, is to examine multi-sensor control strategies in an effort to improve the performance of the controllers, their stability and/or how implementable they are. Up to this point, the primary method of control for systems of this type has been classical, designing single-input, single output controller loops to be closed around each strut. But because of the geometry of the hexapod and the different problems that can occur with some sensors, the classical approach is limited in what it can accomplish. This research shows the benefits to be gained by going to a multiple sensor controller and implementing controllers that are designed using a frequency weighted, linear quadratic gaussian approach. The question of strut failure in a classical controlled system as compared to one using a multi-input, multi-output strategy is also investigated.