基于多维泰勒网的电力系统辨识研究与应用  

Research and Application of Power System Identification Based on Multi-dimensional Taylor Network

作　　者：於晶晶[1] 孙启鸣 YU Jingjing;SUN Qiming(China Railway Shanghai Design Institute Group Co.,Ltd.,Shanghai 200070,China;College of Information Science and Technology,Nanjing Forestry University,Nanjing 210037,China)

机构地区：[1]中铁上海设计院集团有限公司,上海200070 [2]南京林业大学信息科学技术学院,江苏南京210037

出　　处：《江苏海洋大学学报（自然科学版）》2020年第4期63-67,共5页Journal of Jiangsu Ocean University:Natural Science Edition

基　　金：江苏省高等学校自然科学研究资助项目(19KJB510034);南京林业大学教学质量提升工程一流课堂与教学方法改革项目(2019ktjx021)。

摘　　要：系统辨识是非线性系统理论的重要组成部分,根据已知的数据建立系统模型并对内部参数进行在线估计,对电力系统的分析、预报、故障诊断乃至进一步智能化发展都有很大的促进作用。提出一种基于多维泰勒网的新型辨识方法,用来对电力系统进行辨识。多维泰勒网结构简单,对非线性系统具有很强的逼近作用。引入共轭梯度法,采用自适应可变步长调整算法,只需整定内部参数,即可将辨识误差最小化。最后,通过一个数值仿真实例和某35 kV变电所电压运行数据辨识实例,验证了所提出方法的有效性。System identification is an important part of the theory of nonlinear systems.System models established based on known data and online estimation of the internal parameters can greatly facilitate the analysis,forecasting,fault diagnosis,and even further intelligent development of power systems.A novel identification method,which is used to identify the power system,based on multi-dimensional Taylor network,is proposed in this paper.With a simple structure,the multi-dimensional Taylor network has a strong approximation effect on nonlinear systems.Introducing the conjugate gradient method and using an adaptive variable step size adjustment algorithm,only the internal parameters need to be adjusted to minimize the identification error.Finally,a numerical simulation example and a 35 kV substation voltage operation data identification example verify the effectiveness of the method proposed in this paper.

关 键 词：系统辨识 多维泰勒网 电力系统 共轭梯度法 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]