Random Forest Based Real Time Fault Monitoring System for Industries

Inala Vivek Vamsi, Amar Kumar Verma, Sudha Radhika

2018 · 23 citations · 7 references

Concepts

Abstract

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.

References

7