首页|基于神经网络的微电网交错并联变换器的改进自抗扰控制

基于神经网络的微电网交错并联变换器的改进自抗扰控制

扫码查看
针对混合储能微电网在负载突变、扰动加入等多种复杂工况下引起的用电端降压接口的电压质量降低的问题,文章设计了一种以改进线性自抗扰为主要控制器,以BP神经网络为辅助参数优化算法的闭环控制策略.首先,根据六路交错并联变换器的电路拓扑建立时域下的数学模型,并对系统总扰动进行重构,将总扰动分解为模型未知扰动和外界扰动,分别利用观测器进行估计,形成了改进线性自抗扰控制,提高系统的扰动观测能力和观测精度;其次,为使系统获得控制参数的实时优化能力,引入BP神经网络,对控制器参数进行实时整定;最后,搭建了混合微电网六路交错并联变换器的数字仿真模型和半实物仿真模型进行验证,并与PI,LADRC进行了比较.结果显示,所提控制策略不仅具备优异的输出电压质量,而且使系统获得了更好的稳定性与鲁棒性.
Design of a microgrid interleaved parallel converter based on neural networks improved active disturbance rejection control
A disturbance compensation type improved active disturbance rejection control strategy is designed to address the problems of multiple disturbances,large inertia,and long delay in the SCR system of the coal mining machine system.Based on the model information of the SCR system,a mathematical model of the required form of active disturbance rejection was established.A second-order degree of freedom auto disturbance rejection was designed to control it,and the total disturbance was reconstructed to be equivalent to unknown disturbances and external disturbances.A new observer was designed for disturbance compensation,forming a disturbance compensation linear auto disturbance rejection,improving the observer's disturbance observation ability and accuracy.Finally,a digital simulation model of the SCR system is built on the MATLAB/Simulink simulation platform and compared with PI and LADRC.The results show that the disturbance compensation improved active disturbance rejection has better anti-interference and tracking capabilities,verifying the correctness and superiority of the proposed control strategy.

hybrid energy storage microgridvoltage quality degradationimproved linear active disturbance rejection controlfuzzy controlsix way interleaved parallel converter

熊志杰、张大伟、席骊瑭、王彦沣、周哲民

展开 >

国网四川省电力公司, 四川 成都 610041

清华大学 电机工程与应用电子技术系, 北京 100084

混合储能微电网 电压质量下降 改进线性自抗扰控制 模糊控制 六路交错并联变换器

基金委智能电网联合基金重点项目

U2066201

2024

可再生能源
辽宁省能源研究所 中国农村能源行业协会 中国资源综合利用协会可再生能源专委会 中国生物质能技术开发中心 辽宁省太阳能学会

可再生能源

CSTPCD北大核心
影响因子:0.605
ISSN:1671-5292
年,卷(期):2024.42(2)
  • 14