Use this URL to cite or link to this record in EThOS:
Title: Development of a fault detection and diagnosis approach for a binary ice system
Author: Liu, Y.
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2015
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Fault detection and diagnosis (FDD) is an important part to maintain the performance, improve the reliability and prevent energy wastage of the refrigeration systems. Binary ice systems, which have become more commonly employed in both industry and domestic applications, are essentially refrigeration systems using water-ice slurry mixture as a secondary refrigerant. The existence of the ice makes binary ice systems different from conventional liquid chillers, leading to the requirement of a specified FDD method. Therefore, the current research focuses on developing a model based dynamic FDD approach that can capture the unique features of binary ice systems in order to detect some pre-selected faults, including binary ice flow restriction, cooling water flow restriction, incorrect solution concentration, ice generator scraper fault and ice generator motor failure. To provide fault free predictions for the FDD, a dynamic hybrid model of the binary ice system was proposed. The model consisted of an analytical sub-model of the scraped surface ice generator, which was an essential component of the binary ice system that produced ice, and an artificial neural network (ANN) sub-model of the primary refrigeration circuit. The two sub-models were coupled by using two of the ANN model's outputs as the inputs to the analytical model, namely the evaporating temperature and the compressor power consumption, as well as sharing some of the input parameters. The coupled model was validated with data from a 2.5kW laboratory binary ice test rig. The FDD was carried out by monitoring the changes of the residuals of some carefully chosen parameters, using CUmulative SUM (CUSUM) test. Two parameters, namely cooling water temperature difference and evaporating temperature, were monitored for fault detection purpose, and condenser outlet temperature, cooling water temperature difference, discharge temperature and binary ice outlet temperature were observed for fault diagnosis function. An ANN fault classifier was developed to identify the type of the fault by analysing the combinations of the fault diagnosis parameter variations. This FDD method was found to be able to detect and diagnose successfully the pre-selected faults without raising any false alarm, and in addition it was capable of diagnosing three pairs of double fault.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available