Immunity based systems for faulty sensor detection
Conference Paper
Overview
Identity
Additional Document Info
View All
Overview
abstract
Borrowing ideas from biological immune systems, Artificial Immune Systems (AIS's) offer a novel approach to solving many diagnosis and control problems. They provide the advantage of parallel processing (and scalability) with inherent adaptability. This paper details results obtained by applying three general AIS metaphors to the problem of active/online sensor diagnostics. We form an intelligent diagnostic system that can detect erroneous sensors and predict correct values, preventing control and/or diagnostic systems from making incorrect and potentially costly decisions. In more specific terms, the approach proposed herein offers a framework for addressing the problem of detection and correction of faults in sensors as a pre-requisite to the development of robust diagnostic and/or control systems. This paper describes the methodology and preliminary results of a study performed at Texas A&M University on this concept with application to diagnostic modeling of energy management and control systems.
