二阶时变非线性系统的边缘线性化模糊推理建模

本文给出了二阶时变非线性系统的边缘线性化建模方法。该方法根据模糊插值机理将被控对象的模糊推理规则库依时间维度进行划分转化为变系数非线性微分方程。在每个时间间隔内,依据边缘线性化方法,将局部时空结构下的变系数非线性微分方程转化为了常系数线性微分方程,进而迁移到整体时空结构下则成为变系数线性时变微分方程。仿真实验表明,该方法具有较高的逼近精度,是对模糊推理建模的有效推广。

Fuzzy Inference Modeling by Marginal Linearization of Second Order Time-varying Nonlinear System

In this paper, the marginal linearization modeling is employed to the second order time-varyingnonlinear system. Based on the fuzzy interpolation mechanism of fuzzy inference rule base, the nonlinear dif-ferential equations with variable coefficients are established, which use the time dimension to divide into someperiods. In each time interval, based on the marginal linearization method, the local space - time structurenonlinear differential equations with variable coefficients transformed into the constant coefficient linear differential equations. Migrated to the whole space - time structure is linear time - varying differential equation withvariable coefficients. Simulation results show that the method has high approximation accuracy, and it is an effective promotion for the fuzzy inference modeling.