模糊神经网络下的SMES三相五电平变流器控制  

Research on SMES Three-phase Five-level Converter Controlled by Fuzzy Neural Network

作　　者：李俊泓 魏振兴[2] LI Jun-hong;WEI Zhen-xing(Guang’an Vocational Technical College,Guangy an 638000,China;不详)

机构地区：[1]广安职业技术学院,四川广安638000 [2]西南交通大学电气工程学院,四川成都611756

出　　处：《电力电子技术》2021年第3期130-134,共5页Power Electronics

基　　金：国家自然科学基金(61601382)。

摘　　要：可再生能源所存在的不稳定性和间断性对电力系统带来了压力,因而需要设计好的控制算法调节变流装置使超导储能(SMES)系统带动微网实现可靠运行。这里提出了一种复合式模糊神经网络控制策略,首先,研究了基础三相五电平电流型变流器和SMES系统的拓扑结构,并给出被控拓扑模型。进而在载波叠加调制比例积分微分(PID)控制的基础上,提出模糊控制策略与3层BP神经网络结合的控制方法,并完成对比实验分析。结果表明所提方法应用于SMES型微网中所产生的电流纹波和畸变程度较小,总电流变化后调整速率均较快;此外,当微网电压处于不平衡状态时,不容易被谐波所影响。The instability and discontinuity of renewable energy bring pressure to the power system,so it is necessary to design a good control algorithm to adjust the converter and make the superconducting magnetic energy storage(SMES)system driving the micro grid to achieve reliable operation.A hybrid fuzzy neural network control strategy is proposed.Firstly,the topology construction of the basic three-phase five-level current converter and SMES system is studied,and the controlled topology model is given.Then on the basis of carrier superposition modulation proportion integral differential(PID)control,a control method combining fuzzy control strategy and three-layer BP neural network is proposed.A three-phase five-level SMES micro-network system is applied to complete the comparative experimental analysis.The results show that the current ripple and distortion caused by the method applied to SMES micro-grid is small,and the adjustment rate after the total current changes is faster.In addition,when the voltage of the micro grid is in an unbalanced state,it is not easy to be affected by harmonics.

关 键 词：变流器 超导储能 模糊控制 神经网络 

分 类 号：TM46[电气工程—电器]