The Joint Accelerator Conferences Website (JACoW) is an international collaboration that publishes the proceedings of accelerator conferences held around the world.
TY - CONF AU - Kong, Y.B. AU - Hur, M.G. AU - Lee, E.J. AU - Park, J.H. AU - Song, H.S. AU - Yang, S.D. ED - Koscielniak, Shane ED - Satogata, Todd ED - Schaa, Volker RW ED - Thomson, Jana TI - Deep Learning Based Predictive Control for RFT-30 Cyclotron J2 - Proc. of IPAC2018, Vancouver, BC, Canada, April 29-May 4, 2018 C1 - Vancouver, BC, Canada T2 - International Particle Accelerator Conference T3 - 9 LA - english AB - Successful construction of the control system is an important problem in the accelerator. The RFT-30 cyclotron is 30 MeV cyclotron for radioisotope production and fundamental researches. To operate the RFT-30 cyclotron for beam irradiation, the human operators should carefully manipulate the control parameters. If the control does not function properly, it becomes difficult to handle the cyclotron and cannot perform the accurate operations for the control. In this work, we propose a deep learning based model predictive control approach for the RFT-30 cyclotron. The proposed approach is composed of two steps: system identification and a control design. In the system identification procedure, the proposed approach constructs the predictive model of the accelerator using the deep learning approach. In the control design stage, the controller finds the optimal control inputs by solving the optimization problem. To analyze the performance of the proposed approach, we applied the approach into the RFT-30 cyclotron. PB - JACoW Publishing CP - Geneva, Switzerland SP - 2230 EP - 2232 KW - controls KW - cyclotron KW - network KW - simulation KW - operation DA - 2018/06 PY - 2018 SN - 978-3-95450-184-7 DO - 10.18429/JACoW-IPAC2018-WEPAL030 UR - http://jacow.org/ipac2018/papers/wepal030.pdf ER -