液位系统的辅助变量复合建模方法研究  

Research on composite modelling method based on instrumental variable identification for water level system

作　　者：杨彦波[1] 胡婷婷[2] 毛忠国[1] 

机构地区：[1]宁夏电力公司能源科技有限公司,宁夏银川750011 [2]宁夏电力公司电力科学研究院,宁夏银川750011

出　　处：《宁夏电力》2013年第3期50-54,58,共6页Ningxia Electric Power

摘　　要：针对发电过程控制策略设计和计量方法研究过程中,对系统模型要求越来越高的现状,基于辅助变量最小二乘(RIVLS)辨识方法,提出一种分别以一阶惯性加纯滞后模型(FOPDT)和二阶惯性加纯滞后模型(SOPDT)为基本组成部分,以建模误差的均方差为加权因子的复合建模方法,并通过实验室液位系统建模仿真过程,证明新算法的有效性。仿真结果表明:该方法具有提高建模准确性的作用,该建模方法可以扩展应用于其它对模型精度要求较高的生产过程中。The system model is being demanded higher and higher in the control strategies design and measuring method research of the power generating processes,based on the Recursive Instrumen tal Variable Least Square（RIVLS）identification method,puts forward composite modelling method： to take the First Order Process with Deal Time（FOPDT）and Second Order Process with Deal Time（SOPDT）as the basic constituent,and the average square modelling error as the scaled-factor.Finally,combines with the modelling simulation process of lab water level system,proves the effectiveness of the new calculation method.The simulation result shows that the method has the role of improving model ling accuracy,applied in other production process which demanded higher toward the model precision.

关 键 词：辅助变量最小二乘法 加权因子 复合建模 建模误差 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]