基于高阶模糊推理系统的曲线拟合研究

Study on curve fitting based on high-order fuzzy reference system

利用高阶模糊推理系统进行曲线拟合.首先构建高阶模糊推理系统模型,然后基于梯度下降的学习算法对高阶模糊推理系统模型的权值进行改进,最后进行曲线拟合,并与基于最小二乘的曲线拟合结果进行比较.结果表明,高阶模糊推理系统比基于最小二乘的曲线拟合具有更强的非线性处理能力,更适用于曲线拟合.

This paper is mainly focus on curve fitting using a high-order fuzzy reference system. A model of highorder fuzzy reference system is firstly constructed, then an improved learning algorithm founded on gradient descent is used to update the weights of the high-order fuzzy reference system. An example about curve fitting based on high-order fuzzy reference system is given in this paper, the results show that high-order fuzzy reference system has higher capability on nonlinear problem than curve fitting based on least squares, then is more suitable for curve fitting.