基于RLS的汽轮机数字电液调节系统参数辨识  

Recursive least square method based parameter identification of steam turbine digital electric-hydraulic governing system

作　　者：李蔚[1] 张政江[1] 盛德仁[1] 陈坚红[1] 任浩仁[1] 

机构地区：[1]浙江大学机械与能源工程学院,浙江杭州310027

出　　处：《浙江大学学报（工学版）》2008年第10期1761-1764,共4页Journal of Zhejiang University：Engineering Science

基　　金：浙江省教育厅资助项目

摘　　要：采用递推最小二乘法(RLS)进行汽轮机数字电液调节系统的参数辨识,通过双线性变换法将连续模型转化为离散模型,推导了电液转换器和油动机环节的辨识模型.为剔除辨识数据中环节通带外信号的不利影响,采用Butterworth低通滤波器进行滤波处理,根据环节对象的预计通带,设计低通滤波器的通带截止频率和阻带截止频率,有效提高了参数的辨识精度.以某300 MW机组数字电液(DEH)调节系统为研究对象,应用传递函数法建立DEH模型结构图,经该机组负荷扰动实测试验数据的计算验证表明,递推最小二乘法具有辨识速度快、建模效率高特点,适合于复杂的现场对象参数辨识.The recursive least square （RLS） method was employed to identify the parameter. The discrete model was established by employing the dual linear transformation, and the identification models of electric-hydraulic converter and servomotor module were derived. Butterworth low-pass filter was applied to avoid the disadvantage of off-band signal and improve the identification precision. The cut-off frequency of Butterworth low-pass filter was designed according to the forecasting band of the object segment. The transfer function method was used to establish the model of digital electric-hydraulic （DEH） governing system of a 300 MW unit. The disturbed test data of the unit show that the RLS method has a high speed and it is efficient for parameter identification of complicated objects.

关 键 词：递推最小二乘法 参数辨识 数字电液 调节系统 

分 类 号：TK414.3[动力工程及工程热物理—动力机械及工程]