Support vector machine based fault detection approach for RFT-30 cyclotron
详细信息 查看全文
- 作者：Young Bae Kong ; ybkong@kaeri.re.kr" class="auth_mail" title="E-mail the corresponding author ; Eun Je Lee ; Min Goo Hur ; Jeong Hoon Park ; Yong Dae Park ; Seung Dae Yang
- 关键词：Cyclotron ; Fault detection ; Support vector machine ; Principal component analysis
- 刊名：Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
- 出版年：2016
- 出版时间：21 October 2016
- 年：2016
- 卷：834
- 期：Complete
- 页码：143-148
- 全文大小：890 K
文摘
An RFT-30 is a 30 MeV cyclotron used for radioisotope applications and radiopharmaceutical researches. The RFT-30 cyclotron is highly complex and includes many signals for control and monitoring of the system. It is quite difficult to detect and monitor the system failure in real time. Moreover, continuous monitoring of the system is hard and time-consuming work for human operators. In this paper, we propose a support vector machine (SVM) based fault detection approach for the RFT-30 cyclotron. The proposed approach performs SVM learning with training samples to construct the classification model. To compensate the system complexity due to the large-scale accelerator, we utilize the principal component analysis (PCA) for transformation of the original data. After training procedure, the proposed approach detects the system faults in real time. We analyzed the performance of the proposed approach utilizing the experimental data of the RFT-30 cyclotron. The performance results show that the proposed SVM approach can provide an efficient way to control the cyclotron system.