Fault-Tolerant Task Allocation in Networked Control Systems

Abstract

Networked control systems are made fault tolerant in this paper by redistributing the control tasks of the subsystems in case of faults. Active fault tolerance is brought into the overall system by a cooperation of the faulty subsystems and the healthy subsystems rather than by trying to recover the nominal performance of the faulty subsystems. This paper states conditions under which reconfigurable tasks exist and provides a description of the design process of the fault-tolerant task allocation unit that determines the new tasks in dependency upon the fault case.