欢迎访问《核聚变与等离子体物理》官方网站,今天是 分享到:

核工业西南物理研究院 ›› 2023, Vol. 43 ›› Issue (3): 255-262.DOI: 10.16568/j.0254-6086.202303002

• 核聚变工程 • 上一篇    下一篇

CRAFT超导磁体电源控制策略设计

黄荣林1, 2,傅 鹏1, 2,黄连生1, 2,许留伟*1, 2,高 格1, 2,何诗英1   

  1. (1.中国科学院合肥物质科学研究院等离子体物理研究所,合肥 230031;2. 中国科学技术大学,合肥 230026) 
  • 收稿日期:2021-03-25 修回日期:2023-03-15 出版日期:2023-09-15 发布日期:2023-08-31
  • 作者简介:黄荣林(1973-),男,安徽舒城人,博士研究生,从事聚变工程变流电源及控制技术研究。
  • 基金资助:
    国家重大科技基础设施建设项目-聚变堆主机关键系统综合研究设施(2018-000052-73-01-001228) 

Design of control strategy for a superconducting magnet power supply for CRAFT 

HUANG Rong-lin1, 2, FU Peng1, 2, HUANG Lian-sheng1, 2, XU Liu-wei1, 2, GAO Ge1, 2, HE Shi-ying1   

  1. (1. Institute of Plasma Physics, Chinese Academy of Sciences, Hefei 230031; 2. University of Science and Technology of China, Hefei 230026) 
  • Received:2021-03-25 Revised:2023-03-15 Online:2023-09-15 Published:2023-08-31

摘要: 聚变堆主机关键系统综合研究设施(CRAFT)的超导磁体是大惯性负载,其测试电流达到 120kA。超导磁体电源需要并联运行输出功率,在电流反馈控制模式,输出电压快速变化和变流器之间的负荷分配是很大的问题。针对输出电压快速变化问题,设计了基于神经元的 PID 参数自适应的控制方法,减小了系统的超调量;针对变流器并联运行的负荷分配问题,在研究变流器环电流特性的基础上,设计了基于虚拟环流阻抗的均流控制策略,保证了变流器之间负荷的平均分配。最后通过仿真和实验验证了设计方法的可行性。 

关键词: CRAFT;超导磁体;电源;环流控制;神经网络 

Abstract:  The superconducting magnet of Comprehensive Research Facility for Fusion Technology (CRAFT) is a high inertia load with a test current of 120kA. Its power supply provides power in parallel operation, and the output voltage quick-change and the load distribution between converters are the challenge problems based on current feedback control. To solve the output voltage quick-change problem, a PID parameter adaptive control method is proposed on the basis of the characteristics of neural network, thus the overshoot of system is reduced. For the load distribution problem, the share current control strategy based on virtual loop impedance is designed with the study on the circulating current characteristics between converters, which ensures the average distribution of load between the converters. At last, simulations and experiments are conducted, and results validate the feasibility of the proposed methods. 

Key words: CRFFT, Superconducting magnet, Power supply, Circulating current control, Neural network

中图分类号: