A global optimization framework for parameter estimation of a wind generation unit model

摘要

This paper purposes a global optimization method that could be applied in parameter estimation of wind generation unit model. When complex nonlinear models, like a wind generation unit model are used, the parameter estimation based on local optimization methods, such as nonlinear least squares method or Newton's method, may not be able to find parameter values with acceptable accuracy. The global optimization method, named Hyperbolic Cross Point (HCP) method, is proposed to find good initial parameter values that are used as the starting values, to make parameter estimation based on iterative local optimization methods, converge to accurate parameter values. The paper concludes with two case studies which demonstrate application of the HCP in conjunction with local optimization method, where the results obtained are confirmed to be the globally best solution.