2018 · 23 citations · 7 references
ComputingIndustrial EngineeringMachine LearningReal-time SystemsStatistical LearningCondition MonitoringDiagnosisDiagnosticsFault DetectionFault DiagnosisHealth Monitoring SystemsFault ClassificationFault Detection Using Machine LearningMachine Health MonitoringMachine Learning For Fault DetectionMachine State MonitoringRandom Forest ClassificationReal-time Fault DetectionReal-time Monitoring
For increasing machinery availability-condition monitoring is highly desirable so as to reduce consequential damage, and improving efficiency of operation by knowing the traits of machine well in advance. Model-based techniques are efficient for inferring warning and are quite useful in predicting industrial machine internal faults. Such machine faults detected in the earlier stage can prevent more severe damage in due course of time and thereby promising an efficient production in industrial field. However, the conventional methods ought to work with distinct motor, and can not be applied effectively for machines with real time damage or effects. In this paper, the Random Forest (RF) classifier has been used for real-time fault monitoring. The advantage of the proposed method is demonstrated in response to accuracy in prediction and graphical user interface with all its functioning running at the backend to eventually display the condition of the machine.
7
Precise wavelet for current signature in 3ϕϕ IM
Sudha Radhika, G. R. Sabareesh, G Jagadanand et al. · Expert Systems with Applications · 2009 · 46 citations