摘要
提出一种基于 EE(方程误差 )模型的时域辨识算法 ,通过对离散采样数据的多重积分 ( PLPF法 )估计模型参数值 ,直接求取待辨识系统微分方程的系数 ;再根据微分方程的系数 ,由待定系数法得到一个非线性方程组 ,然后用牛顿迭代法求解得出传递函数的模块环节实际参数。该方法的优点在于根据输入输出采样数据直接在时域上进行参数辨识 ,方法简便 ,较好地解决了由传递函数多项式 a,b系数转化为传递函数框图环节参数的实际需求。
A new method for identifying the excitation system parameters of large synchronous generators has been presented. This method,based on EE model,is a tim e domain identification method.Applying multiple integral to discrete sample data, we can get the param eters of the differential equation of the identified system.Newton iteration method is introduced to resolve m odel parameters.According to input and output data,this m ethod identifies parameter in time dom ain directly. Moreover,it has solved the questions that transform the transfer- function coefficient into the traditional m odel param eters. In particular,this m ethod has been applied successfully in a 2 0 0 MW generator.Field tests show thatthis method is reliable.
出处
《电力系统自动化》
EI
CSCD
北大核心
2002年第8期25-28,37,共5页
Automation of Electric Power Systems
