Stochastic Fault Detection...

This entry describes the state-of-the-art and future perspectives on stochastic fault detection, namely, stochastic fault detection and diagnosis (FDD). Both model-based and data-driven FDD methods for stochastic signals and systems have been included, where the use of hypothesis testing, Kalman filtering, system estimation, principal component analysis (PCA), and stochastic distribution control has been discussed for the construction of effective FDD algorithms. Indeed, stochastic FDD constitute an important and integrated part in developing fault-tolerant controls (FTC) for guaranteed safe operation of control systems, of which increased penetration of random factors is inevitable nowadays.