Health monitoring of feedback controlled mechatronic systems

• Main Author: Abdulabbas Gatea Al-Khafaji, Ali
• Published: Cardiff University 2017
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.753537

Health monitoring is essential in guaranteeing the safe, efficient, and correct operation of complex engineering systems. This PhD thesis presents a simulation of a non-linear, experimental-based model of a coupled tank apparatus CE105 under LabVIEW environment. The consideration of a traditional simple tank system is extended via the inclusion of non-linear elements. The simulation is used to accelerate the timescales of the monitoring and controller signals for nominal and faulty behaviour for several operating scenarios. In this study, a detailed simulation with several sources of fault was produced and run with the variety of operating scenarios to study the nominal and faulty behaviour of such mechatronic system. It is concluded that the liquid level will not be affected by fault nature and intensity in the presence of PID controller that covers hidden faults until its signal reaches a certain threshold. Hence, the end of useful life can be predicted by tracking the PID signal at any stage of the operating scenario. Technology advances have impacted upon monitoring, diagnostics and prognostics activities for increasingly sophisticated industrial systems and their operations. In particular, for integrated mechatronic systems, the facility provided by dynamic simulation models in presence of deteriorating faults has been investigated. For informed data-driven prognostic extrapolations, the long-term, time-varying operational profile of the mechatronic system requires recording and analysis. The contribution reported in this study relates to the simulation and experimentally validated, of a CE105 coupled-tank liquid level control system and three individual-thank liquid level system. A Sign Chart Algorithm (SCA) was developed and utilised as a novel controller-based health monitored (CBHM) system. Moreover, from the SCA and the PID signal trend, the remaining useful life of the system has been estimated. Results are reported and discussed for leakage or blockage and pump performance deterioration faults.