An Adaptive Model-Based Approach to the Diagnosis and Prognosis of Rotor-Bearing Unbalance
Open Access
Abstract
Rotating machinery is the fundamental component of almost all industrial frameworks. Therefore, prognostics and health management (PHM) have emerged as crucial requirements for effectively managing and sustaining various systems in a timely manner. The unbalanced fault has been recognized as a significant contributing factor in the development of faults in rotor-bearing systems, eventually causing failure. Thus, it is essential to monitor the unbalance and maintain it within acceptable bounds in order to guarantee the system’s proper operation. Most approaches to the rotor’s unbalance monitoring are model-based instead of data-driven due to the shortage of faulted data. In a derived model-based approach, proper identification of the model’s parameters, e.g., bearing parameters, always plays a very crucial role. Nonetheless, the identified model’s parameters in their initial state would inevitably degenerate during a long-term operation because of aging or environmental changes, such that they are no longer well representative of the real system. In this context, this paper offers an adaptive model-based approach for the assessment of unbalance faults developing over days in a rotor-bearing model. The model is adaptive in the sense that it automatically adjusts its parameters so that they are more closely aligned with the real system. A particle swarm optimization (PSO) scheme is utilized in the parameter identification process. The residual serves as the index for initiating the adaptive process when it is greater than a preset percentage. Individual feature errors work as a gauge to determine which bearing parameters need to be reevaluated. A set of 16-month operational data from a local petrochemical company is used to validate the approach. The unbalanced deterioration trend is evaluated, and results from the adaptive methodology are assessed to show its superiority over the original one. It is also observed that the model’s capacity to anticipate unbalance is greatly enhanced by the adaptive strategy. Finally, future unbalances are explored to show its capacity for continuous monitoring-based maintenance solutions.Keywords
This publication has 33 references indexed in Scilit:
- Dynamic Analysis of Flexible Rotor-Ball Bearings System with Unbalance-Misalignment-Rubbing Coupling FaultsApplied Mechanics and Materials, 2011
- Identification of unbalance in a rotor bearing systemJournal of Sound and Vibration, 2011
- Model based fault diagnosis of a rotor–bearing system for misalignment and unbalance under steady-state conditionJournal of Sound and Vibration, 2009
- Model-based identification of rotating machinesMechanical Systems and Signal Processing, 2009
- Simultaneous estimation of the residual unbalance and bearing dynamic parameters from the experimental data in a rotor-bearing systemMechanism and Machine Theory, 2009
- The use of features selection and nearest neighbors rule for faults diagnostic in induction motorsEngineering Applications of Artificial Intelligence, 2006
- A comparison study of improved Hilbert–Huang transform and wavelet transform: Application to fault diagnosis for rolling bearingMechanical Systems and Signal Processing, 2005
- Identification of Unbalance and Crack Acting Simultaneously in a Rotor System: Modal Expansion versus Reduced Basis Dynamic ExpansionJournal of Vibration and Control, 2005
- Accuracy of Fault Detection in Real Rotating Machinery Using Model Based Diagnostic TechniquesJSME International Journal Series C, 2003
- IDENTIFICATION OF MULTIPLE FAULTS IN ROTOR SYSTEMSJournal of Sound and Vibration, 2002