Multi-Stage Model Predictive Control Algorithm with Eddy Current Compensation for FTPMF
ID:101 View Protection:PUBLIC Updated Time:2023-06-14 01:50:15 Hits:1221 Poster Presentation

Start Time:2023-06-19 09:00(Asia/Shanghai)

Duration:0min

Session:E Poster Session » E2Poster Session 2

Abstract
Many vacuum electronic devices require a power supply to drive magnets to generate a flat-top pulsed magnetic field in their cavities. However, there is a thick copper plating layer in cavities, resulting in severe eddy current hysteresis. In this paper, a control algorithm based on multi-stage model predictive control is proposed for the eddy current compensation of high-power gyrotron terahertz wave sources. Compared to traditional flat-top pulse current sources, the proposed algorithm generates a controlled current overshoot in the magnet, increasing the magnetic field stabilization speed by approximately 50%. The control algorithm has multiple operating states and can achieve computational load balancing by adjusting the sampling time and prediction step length under different states while generating a more stable flat-top pulse current than traditional methods.
Keywords
Model Predictive Controller
Speaker
Zhenglei Wang
Docter Huazhong Univ. of Sci. & Tech.

Shaozhe Zhang

Chunhui Yang

Houxiu Xiao
Huazhong University of Science and Technology

Tonghai Ding

Xiaotao Han
Huazhong University of Science and Technology

Submit Comment
Verify Code Change Another
All Comments
Important Date
  • Conference Date

    Jun 16

    2023

    to

    Jun 19

    2023

  • Jun 15 2023

    Contribution Submission Deadline

  • Jul 02 2023

    Registration deadline

Sponsored By
Huazhong University of Science and Technology, China
(IEEE PELS)
IEEE
Contact Information