Advanced monitoring, fault diagnostics, and maintenance of cryogenic systems

In this Thesis, advanced methods and techniques of monitoring, fault diagnostics, and predictive maintenance for cryogenic processes and systems are described. In particular, in Chapter 1, mainstreams in research on measurement systems for cryogenic processes are reviewed with the aim of dening key current trends and possible future evolutions. Then, in Chapter 2, several innovative methods are proposed. A transducer based on a virtual ow meter is presented for monitoring helium distribution and consumption in cryogenic systems for particle accelerators [1]. Furthermore, a comprehensive metrological analysis of the proposed transducer for verifying the metrological performance and pointing out most critical uncertainty sources is described [2]. A model-based method for fault detection and early-stage isolation, able to work with few records of Frequency Response Function (FRF) on an unfaulty compressor, is then proposed [3]. To enrich the proposal, a distributed diagnostic procedure, based on a micro-genetic algorithm for transducer networks monitoring complex physics systems, is shown [4]. Finally, a GEO algorithm for predictive maintenance scheduling problems, already proposed in our Research Group [5] and compared to GA with encouraging but only preliminary results, is extended to complex plants such as large experimental cryogenic systems. In Chapter 3, the numerical characterization of the proposed methods is reported. Finally, in Chapter 4 the experimental validation for the presented techniques is illustrated.