基于注意力与长短期记忆网络的变压器代理模型

摘要
图/表
参考文献
相关文章 (8)

全文: PDF (1785 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要由于需要考虑换能效率、噪声、体积和质量等因素,电力变压器的设计参数和性能数据往往十分复杂,因此,如何建立变压器代理模型是亟需解决的问题。采用代理模型的优化算法（SBO）能有效解决数值模拟直接优化耗时长的问题。本文用深度学习建立变压器设计参数和性能数据的代理模型,实现变压器性能优化目标的高精度预测,有效降低变压器性能分析与优化所需时间。首先基于长短期记忆网络（LSTM）的深度学习模型,建立非晶合金变压器各个参数间的非线性映射,并加入注意力机制来增强模型的预测效果。最后,通过有限元仿真实验对提出的深度学习代理模型进行验证,并与其他常用的代理模型进行比较,证明了注意力机制与长短期记忆网络代理模型在预测精度方面的优越性。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	金亮
	冯裕霖
	曹佳豪
	王艳阳

关键词 ：非晶合金变压器, 有限元方法, 代理模型, 深度学习, 长短期记忆网络, 注意力

Abstract：The design parameters and performance data of power transformers are often very complex considering many factors such as energy exchange efficiency, noise, volume and weight. Therefore, how to establish transformer surrogate model is an urgent problem to be solved. The surrogate-based optimization (SBO) can effectively solve the problem of long optimization time. In this paper, the surrogate model of transformer design parameters and performance data is established by deep learning to achieve high precision prediction of transformer performance optimization objectives and effectively reduce the time required for transformer performance analysis and optimization. Firstly, based on the deep learning model of long-short term memory network (LSTM), the nonlinear mapping between various parameters of amorphous alloy transformer is established, and the attention mechanism is added to enhance the prediction effect of the model. Finally, the proposed deep learning surrogate model is verified by finite element simulation experiment and compared with other commonly used surrogate models. The results show that the attention and long-short term memory surrogate model is superior in prediction accuracy.

Key words： amorphous alloy transformer finite element method surrogate model deep learning long-short term memory (LSTM) attention

收稿日期: 2020-12-12

基金资助:国家自然科学基金面上项目（51977148）

作者简介: 金亮（1982—）,男,博士,副教授,主要研究方向为工程电磁场与磁技术、电磁场云计算和电磁无损检测等。

引用本文:

金亮, 冯裕霖, 曹佳豪, 王艳阳. 基于注意力与长短期记忆网络的变压器代理模型[J]. 电气技术, 2021, 22(7): 65-71. JIN Liang, FENG Yulin, CAO Jiahao, WANG Yanyang. Transformer surrogate model based on attention and long-short term memory. Electrical Engineering, 2021, 22(7): 65-71.

链接本文:

http://dqjs.cesmedia.cn/CN/Y2021/V22/I7/65

[1] 王辉. 变压器磁场分析法与解析法的应用对比[J]. 变压器, 2000, 37(8): 22-26.
[2] 李冰, 王泽忠, 刘恪, 等. 特高压变压器直流偏磁对绕组电流的影响[J]. 电工技术学报, 2020, 35(7): 1422-1431.
[3] 陈伟根, 滕黎, 刘军, 等. 基于遗传优化支持向量机的变压器绕组热点温度预测模型[J]. 电工技术学报, 2014, 29(1): 44-51.
[4] DANESHMAND S V, HEYDARI H.A diversified multiobjective simulated annealing and genetic algo-rithm for optimizing a three-phase HTS transformer[J]. IEEE Transactions on Applied Superconductivity, 2016, 26(2): 1-10.
[5] 潘再平, 张震, 潘晓弘. 基于QPSO算法的电力变压器优化设计[J]. 电工技术学报, 2013, 28(11): 42-47.
[6] TRIA L R, ZHANG D, FLETCHER J.High-frequency planar transformer parameter estimation using differ-ential evolution[C]//2015 IEEE International Magnetics Conference (INTERMAG), Beijing, 2015.
[7] FORRESTER A I J, KEANE A J. Recent advances in surrogate-based optimization[J]. Progress in Aerospace Sciences, 2009, 45(1): 50-79.
[8] 张剑. 基于代理模型技术的高速列车性能参数设计及优化[D]. 成都: 西南交通大学, 2015.
[9] XIA Bin, REN Ziyan, KOH C S.Utilizing kriging surrogate models for multi-objective robust optimi-zation of electromagnetic devices[J]. IEEE Transa-ctions on Magnetics, 2014, 50(2): 693-696.
[10] 陈亚, 李萍. 基于神经网络的短期电力负荷预测仿真研究[J]. 电气技术, 2017, 18(1): 26-29.
[11] 杨杰, 黄晨, 石恒. 径向基函数神经网络补偿的悬浮球悬浮高度自适应滑模控制[J]. 电气技术, 2020, 21(2): 26-30.
[12] 彭珍瑞, 郑捷, 白钰, 等. 一种基于改进MCMC算法的模型修正方法[J]. 振动与冲击, 2020, 39(4): 236-245.
[13] ACAR E, RAIS-ROHANI M.Ensemble of meta-models with optimized weight factors[J]. Structural and Multidisciplinary Optimization, 2009, 37(3): 279-294.
[14] SHAO Zhenfeng, ZHANG Linjing, WANG Lei.Stacked sparse autoencoder modeling using the synergy of airborne LiDAR and satellite optical and SAR data to map forest above-ground biomass[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2017, 10(12): 5569-5582.
[15] QIU Gao, LIU Youbo, LIU Junyong, et al.Analytic deep learning-based surrogate model for operational planning with dynamic TTC constraints[J]. IEEE Transactions on Power Systems, 2020: 1-8.
[16] BARMADA S, FONTANA N, SANI L, et al.Deep learning and reduced models for fast optimization in electromagnetics[J]. IEEE Transactions on Magnetics, 2020, 56(3): 1-4.