Neural networks for sensor fault correction in structural control

Two neural network architectures are proposed for use in structural control applications: a Failure Detection Neural Network and a Failure Accommodati on Neural Network. The Failure Detection Network monitors structural respon ses and automatically detects sensor failures that can reduce control perfo rmance and effectiveness, while the Failure Accommodation Network accounts for the failed sensors. Together, the networks are a step toward developmen t of an expert diagnostic system for structural applications. Examples of t wo simple structures are used to illustrate the features of the networks. S ensor failures are simulated during control operation, and the ability of t he networks to detect and accommodate the failures is examined. The numeric al results reveal that these networks show promise for automated intelligen t fault detection, identification, classification, and accommodation, and a s such may have potential use in real civil structures. Although the networ ks have been used to detect and account for sensor faults alone, they may a lso be trained for other kinds of failures. Thus, they may have potential f or incorporation into an intelligent structural monitoring system.