标量观测辨识模型参数的自适应调试法

Use of Adaptive Control Method and Structural Adjustment in Parameter Estimation from a Time Series

针对从标量观测辨识模型参数提出自适应结构调试法,适用于标量函数显式依赖于某变量,未知参数显式出现在该变量的演化方程中的情况。通过调试法解决自适应辩识中辩识参数不收敛问题。对Lorenz混沌系统的数值模拟说明本方法的有效性。

An adaptive control method with a structural adjustment is introduced for estimating the unknown parameters when the time series for a scalar function of the variables from a nonlinear dynamical model is given. The method with the structural adjustment is shown to work when the function explicitly depends on a variable with the evolution equation in which the unknown parameter appears explicitly, and the structural adjustment is shown to work should the previously adaptive control method not work. The structural adjustment provides a solution to extract the parameters out from a scalar time series. Numerical simulations are presented for Lorenz chaotic system and Rosslor chaotic system.