Sequential multivariable feedback controller design for fault -tolerant applications
Abstract
A multivariable feedback controller design methodology for fault-tolerant control applications is presented in this dissertation. The class of systems considered are square multivariable systems that have an invertible transfer function matrix. The class of faults addressed includes actuator and sensor failures where the output of the failed device becomes uncorrelated to the input. The proposed controller design methodology is based on a non-diagonal generalization of the sequential loop closure technique. In this technique, the off-diagonal controllers are designed during the sequential design process to achieve the fault-tolerant objectives. Two categories of fault-tolerant control are addressed: passive and active fault-tolerance. For passive fault-tolerant control applications, a time-invariant feedback controller is designed which retains closed-loop stability and limp-home performance despite anticipated actuator and sensor failures. For active fault-tolerant control applications, a different feedback controller matrix is specifically designed for each anticipated failure scenario. Fault detection and isolation (FDI) techniques are used to determine when a failure has occurred. Upon determination of a fault, the feedback controller is then switched to the pre-computed control law that corresponds to the identified failure scenario.
Degree
Ph.D.
Advisors
Franchek, Purdue University.
Subject Area
Mechanical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.