超导磁储能系统的序贯克里金优化方法

超导磁储能系统(superconducting magnetic energy storage，SMES)是超导应用研究的热点。SMES利用超导磁体的低损耗和快速响应能力，通过电力电子型变流器与电力系统相连，组合为一种既能为其储存电能又能为其释放电能的多功能电磁系统。SMES的先进功能主要体现于，它能大容量超低损耗的储存电能、改善供电质量、提高系统的稳定性和可靠性。该文以SMES的优化设计(IEEE TEAM Workshop Problem 22)为例，介绍了序贯优化方法和克里金(Kriging)统计近似模型在低维和高维、离散域和连续域优化问题中的应用。优化结果显示，该优化方法能在保证设计精度的前提下，极大降低有限元的计算量。如3参数优化问题中有限元的计算量比直接优化的1/10还要少；而8参数优化问题中有限元的计算量约为直接优化的1/3。从而该方法可广泛应用于电磁装置的优化设计问题。

Sequential Kriging Optimization Method for Superconducting Magnetic Energy Storage

Superconducting magnetic energy storage (SMES) is a research hotspot in the application field of superconducting materials. Using the property of low loss and fast response of superconducting magnet, SMES is employed as a multifunctional electromagnetic system to store and release the power for power system with the connection of power electronic converters. SMES has capacity to reach the objects, such as storing large amount of energy with very low loss, improving the power supply quality and enhancing the stability and reliability of power system. IEEE TEAM Workshop Problem 22 deals with the optimization design of SMES, which is a benchmark problem including low and high dimensional, discrete and continuous optimization cases. To address this problem, sequential optimization method (SOM) and Kriging model are fully discussed. It can be seen that SOM can obtain satisfactory solutions, and the overall computational effort needed is much less than that by direct optimization method. For the 3 parameters problem, the number of finite element sample points is less than 1/10 compared with that of direct method; and for the 8 parameters problem, that number is about 1/3 compared with that of direct method. So the proposed method can be widely employed in the optimization design problem of electromagnetic devices.