|
| 摘要: |
| 提出了用一系列简单函数分段或分区描述大幅度电压变化情况下负荷、变结构负荷和复杂非线性负荷特性的样条函数模型。介绍了一元样条函数模型参数辨识的样条最小二乘法,提出了确定样条函数模型中多项式系数和内节点数值的参数辨识的2层优化算法,其中内层优化问题为内节点数量和数值确定情况下的多项式系数的辨识,外层优化问题为内节点数值的辨识,问题求解采用步长加速法。通过改变样条函数模型中多项式的阶次、内节点数目和内节点数值,可用样条函数模型灵活方便地描述许多复杂的电力负荷特性,并且具有传统数学模型的简单结构和形式,为模型结构或形式未知情况下的负荷建模提供了新的途径。仿真建模实例证明了所提模型与算法的有效性。 |
| 关键词: 负荷建模,样条函数,参数辨识 |
| DOI： |
| 分类号:TM714 |
| 基金项目: |
|
| Electric load models described by spline function |
|
ZHANG Jian SHEN Feng HE Ren-mu
|
| Abstract: |
| The spline function model,which can be represented in separate sections or areas by a series of simple functions,is proposed to describe the characteristics of the load under large voltage variation,the variable-structure load and the complex nonlinear load.The least square method used for the spline function model parameter identification is introduced.The two-layer optimization algorithm is studied to determine the coefficients of polynomial and the values of inside nodes in the spline function model.The inner layer optimization is the coefficient identification of polynomial while the number and locations of inside nodes are determined.The outer layer optimization is the identification of inside node number and locations.The direct optimization algorithm,Hooke-Jeeves method,is used in the parameters identification.By changing the orders of polynomial in spline function,the number of inside nodes and their locations,the model can describe the characteristics of complex load conveniently with the same simple structure as the classical mathematic model.It provides a new effective method for load modeling with model structure unknown.Simulation examples prove the validity of the proposed method. |
| Key words: load modeling,spline function,parameter identification |
