Equipment Failure Prediction Research Articles

Proposal of fault prediction system for underground installations

The absence of intelligent systems for fault prediction in components and equipment in underground installations does not allow a correct structure for maintenance programs to reduce intervening maintenance actions. Thus, it is necessary to devise a computer system that can conjugate potential causes and factors of failure as input data that can correctly inform output data, which represents a component or equipment failure prediction. The aim of this paper is to present a proposal system to support maintenance analyses, where sensors of temperature, water level and vibration, can provide real time data to analyze the isolation condition of transformers and switches within underground chambers. The paper also presents a new use of a neural network to alarm partial discharges (PD) activities. A case is presented where hardware and software are developed and an underground chamber is used as a real test case. In this research, cost, performance, environmental hostility, and the availability of installed equipment in the Brazilian market were considered. After the methodology was verified, the cost was compared with imported equipment, and the result was approximately a 50% cost reduction. Issues of telemetry, hub specification, and sensor installation were also evaluated.

Fuzzy and grey theories in failure mode and effect analysis for tanker equipment failure prediction

Failure Mode and Effects Analysis (FMEA) has been extensively used for examining potential failures in many industries. Unlike the traditional FMEA, the risk factors of occurrence (O), severity (S) and detection (D) of each failure mode are evaluated by linguistic terms and fuzzy ratings in this research, and their relative weights are considered, which make the results more useful and practical. Both grey theory and fuzzy theory are applied in FMEA, in which fuzzy set theory is used to calculate the fuzzy risk priority numbers (FRPNs), and grey theory is applied to calculate the grey relational coefficient. The rankings of the failure modes are both determined by FRPNs and the grey theory. An example of oil tanker equipment failure is provided, and the results show that the evaluation of failure modes by both fuzzy theory and grey theory are quite similar. The practical application of fuzzy and grey theories proposed in this paper helps to enhance the reliability of the prediction, and the predicted ranking of equipment failure can be used for better decision-making concerning inspection and maintenance, which in turn will make tanker shipping more safe and reliable.

하둡 기반의 통합설비 모니터링시스템 설계 및 구현 사례 연구

SCADA and DCS that have performed automatic control and monitoring activities increase the productivity of enterprise in industries. In such systems, although their performance had been improved, there are still many deficiencies in predictive maintenance which can foresee the risk of any kinds of accidents. Because the data acquisition systems of main facilities are being distributed throughout the whole plant and therefore, integration of data obtained from the systems is very difficult. Accordingly, techniques that acquire meaningful information from the gathered data through realtime analysis still need to be improved. This paper introduces a developed facility monitoring system which can predict equipment failure and diagnose facility status through big data analysis to improve equipment efficiency and prevent safety accidents.

SU-E-T-141: Evening Output Check Improves Daily QA Reliability

Purpose: To investigate the benefits of performing a routine evening output check during IMRT QA. Having a secondary dosimetrically accurate measurement at the end of the treatment day proves extremely useful when deterioration of machine output is inferred from the morning QA. Methods: A two-dimensional ionization chamber array (MatriXX) is used for IMRT QA on six linear accelerators. Out of the 12 beams studied six are 6MV, while the rest are 10MV. As part of our IMRT QA protocol, a 10×10 100MU field is routinely recorded as a way to correct for machine output variations. The central axis dose is stored in a database and compared with the morning QA database over a period of 18 months. The distributions of central axis readings are analyzed in terms of range and relative standard deviation. Results: The standard deviation of the readings (averaged over all 12 beams) in the morning QA measurements is 0.88%, while the evening variability is reduced to 0.68%. The range of the evening data (the difference between the lowest and highest reading) is 40% less than the spread in the morning readings signifying stabilization of the machine output during the course of the treatment day. In the case of one treatment machine the morning QA readings of both 6MV and 10MV showed a downward trend not observed in the evening data — a warning sign predicting morning QA equipment failure. Conclusion: The additional beam output information routinely collected in the evening during IMRT QA shows reduced variability compared to the data obtained by the morning QA device. Additionally, the trend of evening output can be compared with the daily QA results to predict equipment failure, and can be used to guide the decision making when the morning readings exceed the predetermined warning limits.

Application of Improved BP Network in Failure Forecasting

An improved BP neural network prediction method is used for collecting pipe equipment failure prediction and comparing with the improved BP neural network in front, which demonstrates that the improved BP neural network algorithm to the collecting pipe failures has better predictive power.

Complex Equipment Failure Prediction Technology Based on Virtual Environment

Complex equipment used in the process, due to changes in running load, vibration, shock and other incidental factors and other confounding factors can lead to changes in equipment and technology status, fault diagnosis of the current state of change brought about by these technologies use is not enough information, relying solely on traditional sensor technology can not test some special devices for signal extraction and data acquisition features. Therefore, the establishment of complex equipment and more operating conditions of knowledge acquisition simulation environment, considering the failure of complex equipment failure caused by the internal laws and mechanisms, to the essence of the phenomenon of failure analysis, combined with the physical prototype test results and raw statistical data, research the actual process of using complex equipment, considering the state of properties, condition monitoring parameters and the integration of traditional knowledge, expertise and reasoning techniques to be applied to physical prototype of a certain conclusion, not only for the early prediction of potential failures, laid the theoretical foundation for forecasting , and fault prediction system for the complex equipment, the development provides important data and technical support to solve the complex equipment maintenance and support tasks needed problem.

The Vibration Fault Diagnosis of High-Speed Rotating Machinery Based on the Simulation

Many researchers explore the issue about how to predict equipment failures and how to analyze the cause of the failures and to solve them. The mass imbalance is the most common cause of the malfunction of the rotating machinery. Based on the theory of rotor dynamics, we can diagnose the failure and its causes more accurately by applying the software to doing simulation analysis of characteristics of rotor dynamic. And, it provides the reliable basis for the further solutions of the failures. In this paper, we analyze a corporation's compressor with abnormal vibration in operation, and finding out the reasons for its vibration. According to the calculated results, we transformed the local structures of its rotor. After transformation, the compressor runs smoothly.

Using GenRel for reliability assessment of mining equipment

PurposeThe purpose of this paper is to formulate, develop and test a reliability assessment model (GenRel) based on genetic algorithms.Design/methodology/approachUsing genetic algorithm based modelling technique, a computer model was developed to predict mine equipment failures from historical data. Two different approaches in application of this technique are demonstrated.FindingsA case study representing a test for convergence of the model was successfully performed. This is an indicator that GenRel can be used to predict equipment failures using a genetic algorithm based modeling technique.Practical implicationsThe use of classical statistical techniques has proven to be an effective tool for reliability analysis of mining equipment. This paper presents an efficient alternative to these classical probability based reliability analysis methods. GenRel is a software solution which performs predictive reliability based upon genetic algorithms (GAs). The advantage of using this technique is the fact that the assumptions based on GAs are much simpler compared to classical statistical methods. The computer model is developed to accept a variety of user input data, most importantly, the ability to use real life historical data in the form of Time Between Failures (TBFs) or Time To Repair (TTRs).Originality/valueThe proposed research offers an alternative method to conventional statistically based reliability analysis and may lead to the foundation of a new approach for reliability assessment with potential applications in other industrial fields as well.

Oil analysis cost‐effective machine condition monitoring technique

This paper describes the development of the use of lubricant analytical programmes and trend analysis to optimise oil change intervals and to predict equipment failure. The various analytical methods are covered, as are the most frequently occurring lubricant applications where such condition monitoring programmes are most appropriate.