Auxiliary signal design for failure detection in uncertain systems

An auxiliary signal is an input signal that enhances the identi- fiability of a model based on input-output observations. Assuming that the normal and the failed behaviors of a process can be modeled by two linear uncertain systems, failure detectability can be seen as a multi-model iden- tification problem. In this paper, we extend previous results on auxiliary signal design for multi-model identification to a larger class of uncertain systems. There are two approaches to the problem of failure detection and isolation. The firstis a passive approach where the detector monitors the inputs and the outputs of the system and decides whether (and if possible what kind of) a failure has occurred. This is done by comparing the measured input-output behavior with "normal" behavior of the system. The passive approach is used to continuously monitor the system in particular when the detector has no way of acting upon the system, for material or security reasons. Most of the work in the area of failure detec- tion is geared towards this type of approach (1), (12), (15). The active approach to failure detection consists in acting upon the system on a periodic basis or at critical times using a test signal, which we call an auxiliary signal, in order to exhibit abnormal behaviors which would otherwise remain undetected during normal operation. The detector can act by taking over part or all of the inputs of the system for a period of time: the test period. The decision whether or not the system has failed should be made at (and if possible before) the end of the test period. The structure of the failure detection method considered in this paper is depicted in Figure 1. noise detection signal