Study of on-line condition monitoring and fault feature extraction for marine diesel engines based on tribological information

Abstract

Reliability and safety issues on marine diesel engines have been received and still need considerable attentions. The literature review indicates that a large amount of failures are caused by abnormal wear of the diesel engine components. It is therefore essential to monitor the engine condition using the tribological information. To further promote the oil monitoring technology into industrial application, a new on-line condition monitoring and remote fault diagnosis system for marine diesel engines is proposed in this article. The new system consists of an on-line tribological signal acquisition model in the ship, a remote feature extraction model and a fault diagnosis model in the laboratory center. The thirdgeneration telecommunication (3G/B3G) which is called wireless communication system is adopted to connect the remote ship with laboratory center. Nine wear characteristics are extracted to detect the engine faults. Experimental tests are implemented on a diesel engine in a real ship named ''Changjing 2'' to evaluate the performance of the proposed fault diagnosis system. The results show that index of particle covered area can be treated as the best feature for marine diesel engines fault detection. The results also reflect that the new system offers satisfactory on-line fault diagnosis ability and is effective for the diesel engine fault diagnosis in practice.

Remote fault diagnostic model for tribological systems in marine diesel engine with two-level self-organizing map network

“…The online ferrographic sensor samples the wear debris image from the oil, and IPCA is calculated with the image digital processing technology [10]. Among the oil parameters collected by this system, Yan and Li in [11] find that IPCA and oil viscosity are the two parameters reflecting the oil condition best.…”

Development and trend of condition monitoring and fault diagnosis of multi-sensors information fusion for rolling bearings: a review

“…Experimental results showed that the fusion technology has a higher reliability and applicability compared to the single detection technology. Ebersbach et al [140] combined vibration with oil analysis and wear debris analysis to establish an expert system to realize the condition monitoring of rotating machinery. Tests results showed that the expert system has a higher reliability and universality compared to manual experience to judge the operation status of the machine.…”

Lubricating Oil Remaining Useful Life Prediction Using Multi-Output Gaussian Process Regression

“…To improve system reliability and prolong remaining useful life (RUL), predictive maintenance (PdM) or condition-based maintenance (CBM) strategies rely on efficient condition monitoring (CM). Among numerous online and offline CM technologies [1][2][3][4], lubricant condition monitoring (LCM) provides early-stage warning ability and multiple other indicators to define the system state [5][6][7]. Wear debris concentration (WDC) is considered as an indicator that can closely define both the lubricant and system state.…”

“…Several other model-based approaches [13][14][15], statistical approaches [16][17][18][19] and hybrid approaches [20], [21] have been widely adopted for degradation trend analysis. These approaches pose certain limitations: (1) For model-based approaches, accurate mathematical modeling of the complex system with non-linearity and uncertainty is difficult; (2) Statistical approaches are single feature focused and not able to pick patterns and auto-state observations, and tend to be harder for modeling of non-monotonic degradation trends; (3) For hybrid methods, modeling of the complex system considering the environmental and external factors still remains a challenge.…”