Reconfigurable fault-tolerant control system for a segmented telescope testbed

Abstract

The design of fault-tolerant control systems is one of the leading issues for highly critical systems. In fact, fault-tolerant control system is used to improve system reliability after failure, maintainability which decreases maintenance time and survivability which is critical in preventing damage. Most control systems are designed to meet performance requirements in the presence of modeling errors, noise, and disturbances. Yet in the presence of faults these robust control systems can not guarantee closed-loop stability or performance. Therefore, a class of control systems called active fault-tolerant control systems that can tolerate various types of faults at the same time maintain acceptable level of performance is considered. This paper presents the research conducted at the NASA sponsored Structures, Pointing and Control Engineering (SPACE) laboratory in designing a decentralized reconfigurable control (DRC) system for a segmented reflector testbed. The objective of the DRC system is to develop controllers that satisfy performance requirements for both normal and/or anomalous (failure) situations. The objective is accomplished by integrating a nominal controller, sensor fault detection and isolation (SFDI), and reconfigurable controllers. The number of computational operations needed to implement a reconfigurable control system for large-scale systems is very large; therefore a centralized reconfigurable control system is usually not feasible. However, with a decentralized approach the total task of the reconfigurable control system is divided into a number of relatively small tasks, which can be implemented by a number of processors working in parallel. The objective of this research is broader in the sense that the methods and tools developed in this study shall be useful for and applicable to a wide variety of reconfigurable control applications.