Diagnostic monitoring of dynamic systems using artificial immune systems

[摘要] The natural immune system is an exceptional pattern recognition system based onmemory and learning that is capable of detecting both known and unknownpathogens. Artificial immune systems (AIS) employ some of the functionalities of thenatural immune system in detecting change in dynamic process systems. Theemerging field of artificial immune systems has enormous potential in the applicationof fault detection systems in process engineering.This thesis aims to firstly familiarise the reader with the various current methods inthe field of fault detection and identification. Secondly, the notion of artificial immunesystems is to be introduced and explained. Finally, this thesis aims to investigate theperformance of AIS on data gathered from simulated case studies both with andwithout noise.Three different methods of generating detectors are used to monitor various differentprocesses for anomalous events. These are:(1) Random Generation of detectors,(2) Convex Hulls,(3) The Hypercube Vertex Approach.It is found that random generation provides a reasonable rate of detection, whileconvex hulls fail to achieve the required objectives. The hypercube vertex methodachieved the highest detection rate and lowest false alarm rate in all case studies.The hypercube vertex method originates from this project and is the recommendedmethod for use with all real valued systems, with a small number of variables at least.It is found that, in some cases AIS are capable of perfect classification, where 100%of anomalous events are identified and no false alarms are generated. Noise has,expectedly so, some effect on the detection capability on all case studies. Thecomputational cost of the various methods is compared, which concluded that thehypercube vertex method had a higher cost than other methods researched. Thisincreased computational cost is however not exceeding reasonable confinestherefore the hypercube vertex method nonetheless remains the chosen method.The thesis concludes with considering AIS's performance in the comparative criteriafor diagnostic methods. It is found that AIS compare well to current methods and thatsome of their limitations are indeed solved and their abilities surpassed in certaincases. Recommendations are made to future study in the field of AIS. Further theuse of the Hypercube Vertex method is highly recommended in real valued scenariossuch as Process Engineering.

[发布机构] Stellenbosch University