Master thesis and internship[BR]- Master's thesis : LQG controller development for active isolation of future gravitational wave detectors from low-frequency seismic motion[BR]- Integration Internship

Abstract

The mysteries of the universe have always fascinated people on earth. Hence, methods to observe and study it in details have been and are still being developed. Since the first detection of gravitational wave at LIGO (Laser Interferometer Gravitational-wave Obser- vatory) in 2015, a new tool to study the universe was added to the panel of existing ones. It is called gravitational wave astronomy. However, the detection of such waves is a technical challenge as their effects are ex- tremely small. As a result, the detectors that aims to capture them must have a tremen- dous degree of precision. The various noise sources to which they are subjected must then be drastically reduced. An important one is the vibration of the detector due to seismic motion. Passive as well as active isolation systems are thus employed to isolate the detector from the ground motion. The goal of this work is to design a type of optimal controller called linear quadratic Gaussian (LQG) controller on an existing experimental isolation platform. The design is separated into two main steps. The first is the development of a full-state observer to estimate the states of the isolation platform from its outputs. The second is the design of a linear quadratic regulator (LQR) to control the system. The two are then combined to form the LQG controller to be used as feedback in a closed-loop system with the initial plant. The performances of this controller are finally analysed and discussed. From this analysis, it emerged that the controller allows to increase the isolation performances of the platform by about one order of magnitude between 0.1 Hz and 1 Hz and by two order of magnitudes from 1 Hz to 10 Hz. Therefore, conclusion has been made that this type of controller were appropriate to provide good isolation performances to the experimental platform in the control bandwidth [0.1,10] Hz. The next step is now to implement it experimentally to address its performances in real situations.