| Title |
Comparison of Fault Detection Technique of Pump System for AI based Smart EOCR |
| Authors |
이경민(Kyung-Min Lee) ; 박철원(Chul-Won Park) |
| DOI |
https://doi.org/10.5370/KIEE.2023.72.9.1012 |
| Keywords |
AI; DNN; EOCR; Fault Detection; Fault Prediction; MCC; Pump System; SVM |
| Abstract |
If operation is interrupted due to unwanted motor fault by various reasons, it can cause enormous damage, including recovery time, cost, and quality loss. To avoid this problem, an EOCR (Electronic Over Current Relay) that monitors and protects the motor is being widely used. Recently, the concept of fault prediction based on the prognosis according to the state is attracting attention. In this paper, in order to improve the smart EOCR for MCC (Motor Control Center) with fault prediction, we detect and compare fault of pump systems using AI (Artificial Intelligence) technology. First, a pump system for smart EOCR is introduced, and a data set is constructed by collecting normal and abnormal state data of pump system. We design and implement a fault detection technique based on DNN (Deep Neural Network) and SVM (Support Vector Machine). Finally, by comparing the simulation results of the two fault detection techniques using AI, the performance evaluation is carried out. |