Application of non-equal interval GM(1,1)model in oil monitoring of internal combustion engine

The basic difference non-equal interval model GM(1,1) in grey theory was used to fit and forecast data series with non-equal lengths and different inertias, acquired from oil monitoring of internal combustion engines. The fitted and forecasted results show that the length or inertia of a sequence affects its precision very much, i.e. the bigger the inertia of a sequence is, or the shorter the length of a series is, the less the errors of fitted and forecasted results are. Based on the research results, it is suggested that short series should be applied to be fitted and forecasted; for longer series, the newer datum should be applied instead of the older datum to be analyzed by non- equalinterval GM(1,1) to improve the forecasted and fitted precision, and that data sequence should be verified to satisfy the conditions of grey forecasting.

Study on Lubricating Oil Monitoring Technology

Lubricating oil monitoring has been proven to be an effective method for detecting and diagnosing machinery failures and essential for realizing condition based maintenance. In this paper, mathematical statistics methods for determining the oil parameters featuring machinery failures and the parameters＇ probability distribution functions and their thresholds are put forward.

OIL MONITORING DIAGNOSTIC CRITERIONS BASED ON MAXIMUM ENTROPY PRINCIPLE

A method of applying maximum entropy probability density estimation approachto constituting diagnostic criterions of oil monitoring data is presented. The method promotes theprecision of diagnostic criterions for evaluating the wear state of mechanical facilities, andjudging abnormal data. According to the critical boundary points defined, a new measure onmonitoring wear state and identifying probable wear faults can be got. The method can be applied tospectrometric analysis and direct reading ferrographic analysis. On the basis of the analysis anddiscussion of two examples of 8NVD48A-2U diesel engines, the practicality is proved to be aneffective method in oil monitoring.

Oil monitoring methods based on information theory

To evaluate the wear condition of machines accurately,oil spectrographic entropy,mutual information and ICA analysis methods based on information theory are presented. A full-scale diagnosis utilizing all channels of spectrographic analysis can be obtained. By measuring the complexity and correlativity,the characteristics of wear condition of machines can be shown clearly. The diagnostic quality is improved. The analysis processes of these monitoring methods are given through the explanation of examples. The availability of these methods is validated and further research fields are demonstrated.

Progress and trend of sensor technology for on-line oil monitoring

Oil monitoring constitutes an important and essential component of condition monitoring technologies and has distinguished advantages in revealing wear,lubrication and friction conditions of tribo-pairs.Newly developed on-line/in-line oil monitoring technologies extend the merits into real-time applications and demonstrate significant benefits in maintenance and management of equipment.This paper comprehensively reviews the progress of on-line/in-line oil monitoring techniques including sensor technologies,their scopes and industrial applications.Based on the existing developments and applications of the sensors for oil quality and wear debris measurements,the trends for future sensor developments are discussed with focuses on accurate,integrated and intelligent features along with exploring a fundamental issue,that is,acquiring the knowledge on degradation mechanisms which has not received sufficient attention until now.Current status of applications of on-line oil monitoring is also reviewed.Although limited reports have been found on this topic,increasing awareness and encouraging progress in on-line monitoring techniques are recognized in applications such as aircraft,shipping,railway,mining,etc.Key fundamental issues for further extending the on-line oil monitoring techniques in industries are proposed and they include long-term reliability of sensors in harsh conditions,and agreement with fault or maintenance determination.

Research on the Correlation Between Oil Menitoring and Vibration Monitoring in Information Collecting and Processing Monitoring

Oil monitoring and vibration monitoring are two principal techniques for mechanical fault diagnosis and condition monitoring at present. They monitor the mechanical condition by different approaches, nevertheless, oil and vibration monitoring are related in information collecting and processing. In the same mechanical system, the information obtained from the same information source can be described with the same expression form. The expressions are constituted of a structure matrix, a relative matrix and a system matrix. For oil and vibration monitoring, the information source is correlation and the collection is independent and complementary. And oil monitoring and vibration monitoring have the same process method when they yield their information. This research has provided a reasonable and useful approach to combine oil monitoring and vibration monitoring.

Information Fusion of Online Oil Monitoring System Using Multiple Sensors

Machine lubrication contains abundant information on the equipment operation. Nowadays, most measuring methods are based on offline sampling or on online measuring with a single sensor. An online oil monitoring system with multiple sensors was designed. The measurement data was processed with a fuzzy intelligence system. Information from integrated sensors in an oil online monitoring system was evaluated using fuzzy logic. The analyses show that the multiple sensors evaluation results are more reliable than online monitoring systems with single sensors.

Research on mechanical wear life feature fusion prediction method based on temporal pattern attention mechanism

In order to solve the problem of low prediction accuracy when only vibration or oil signal is used to predict the remaining life of gear wear,a gear wear life feature fusion prediction method based on temporal pattern attention mechanism is proposed.Firstly,deep residual shrinkage network(DRSN)is used to extract the features of the original vibration time series signals with low signal-tonoise ratio,and the vibration features associated with gear wear evolution are obtained.Secondly,the extracted vibration features and the oil monitoring data that can intuitively reflect the wear process information are jointly input into the bi-directional long short-term memory neural network based on temporal pattern attention mechanism(TPA-BiLSTM),the complex nonlinear relationship between vibration features,oil features and gear wear process evolution is further explored to improve the prediction accuracy.The gear life cycle dynamic response and wear process signals are obtained based on the gear numerical simulation model,and the feasibility of the proposed method is verified.Finally,the proposed method is applied to the residual life prediction of gear on a test bench,and the comparison between different methods proved the validity of the proposed method.

Decision Support System for Maintenance Management Using Bayesian Networks

The maintenance process has undergone several major developments that have led to proactive considerations and the transformation fiom the traditional ＂fail and fix＂ practice into the ＂predict and prevent＂ proactive maintenance methodology. The anticipation action, which characterizes this proactive maintenance strategy is mainly based on monitoring, diagnosis, prognosis and decision-making modules. Oil monitoring is a key component of a successful condition monitoring program. It can be used as a proactive tool to identify the wear modes of rubbing pans and diagnoses the faults in machinery. But diagnosis relying on oil analysis technology must deal with uncertain knowledge and fuzzy input data. Besides other methods, Bayesian Networks have been extensively applied to fault diagnosis with the advantages of uncertainty inference; however, in the area of oil monitoring, it is a new field. This paper presents an integrated Bayesian network based decision support for maintenance of diesel engines.

FTIR analysis and monitoring of used synthetic oils operated under similar driving conditions

The processes of degradation of engine oils operated in passenger cars of a uniform fleet of 25 vehicles were analyzed for oxidation content using infrared (IR) spectroscopy. As part of the experiment, the changes in engine oils occurring during actual operation (under conditions which can be described as "harsh", i.e., short distance driving, frequent starting of the engine, and extended engine idling) have been observed. An evaluation of the Fourier transform infrared spectroscopy (FTIR) spectrum of an engine oil sample was presented. The infrared spectra of both fresh and used oils were recorded with the Thermo Nicolett IS5. The tests were conducted according to the Appendix A2 of ASTM 2412. For the used engine oil differentiation process, FTIR spectra were analyzed in the regions of 1,700–2,000 cm-1 and 3,600–3,700 cm-1. The FTIR spectrometry is demonstrated to be effective for the analysis and monitoring of processes of oxidation and shown to provide rapid and accurate information relating to the aging process of engine oils. The results may facilitate decision-making regarding the service life of engine oils. The achieved dependencies can make it possible to upgrade the sensor assembly consisting of an FTIR source.