摘要详情

ID / 提交时间

244 / 2024-06-30 16:17:31

标题

New Fast Switch and Dynamic Control Method for Data Centers

关键字

Constant air-gap magnetic actuator, Fast switch , PSO-RBF, Displacement prediction

主题及专题

6)Digital design technologies (CAE) in switchgears

状态

全文录用

作者

li qingshen / Dalian university of technology

Dong Enyuan / Dalian Uni. of Tech.

WANG Yongxing / Dalian Uni. of Tech.

Wang Gongrun / State Grid Shandong Electric Power Research Institute

Liyan Zhang / School of Electrical Engineering； Dalian University of Technology

Tong Bo / Dalian University of Technology

Zhou Jian / Dalian University of Technology

Debin Huang / Dalian University of Technology

摘要

Data centers often adopt a dual power supply architecture to ensure reliability, with the core being the fast switch. In the event of a fault in the main circuit, the switch is required to respond within 5 milliseconds and switch to the backup power supply However, the switch have fast speeds excessively which can lead to the main circuit arc not being fully extinguished before the switch is made to the backup power supply circuit, potentially causing an electrical short circuit between the dual circuits and triggering a power supply accident. Therefore, this paper proposes a fast switch based on a constant air-gap magnetic actuator. The actuator used by this switch is characterized by its flexible response and rapid initiation, which can meet the requirements of the fast switch for operation time. However, in the scenario of dual-circuit power supply, the fast speed of the fast switch can cause arc failed extinguish in the process of switching. Thus, further research on displacement control technology is needed to improve the reliability for its switching action. This study first proposes a fast switch based on a constant air-gap magnetic mechanism and explains its basic working principle; then, it investigates a position prediction method based on PSO-RBF (Particle Swarm Optimization - Radial Basis Function). On this basis, a displacement prediction model for the constant air-gap magnetic actuator is constructed using finite element simulation technology. Finally, the effectiveness of the PSO-RBF position prediction model is verified through simulation, and a piecewise PWM (Pulse Width Modulation) strategy is executed based on the predicted displacement curve. The results show that the method can effectively control the motion speed of the mechanism.