基于模糊推理系统的非线性组合建模与预测方法研究

基于模糊推理系统在紧支集中能够逼近任意非线性连续函数的特性，提出了一种基于Takagi-sugeno模糊规则基的非线性组合建模与预测新方法，以克服线性组合预测方法在解决非平衡时间序列组合建模问题所遇到的困难和存在的不足，并给出了相应的基于学习自动机层次结构的优化算法确定模糊系统的参数和模糊子集的划分，理论分析和大量的经济预测实例表明：该方法具有很强的学习与泛化能力，在处理诸如经济时间序列这种具有一定程度不确定性的非线性系统组合建模与预测方法有很好的应用。

Research on the Technique of Nonlinear Combination Modeling and Forecasting Based on Fuzzy Inference System

Based on the property that fuzzy inference system can uniformly approximate any nonlinear multivariable continuous function arbitrarily well, a new nonlinear combination forecasting method is presented to overcome the difficulties and drawbacks in combined modeling non stationary time series by using linear combination forecasting method. Furthermore, the optimization algorithm based on a hierarchical structure of learning automata is used to identify the membership functions in the antecedent part and the real numbers in consequent part of the inference rule. Theoretical analysis and forecasting results related to numerical examples all show that the new technique has reinforcement learning properties and universalized capabilities. With respect to combined modeling and forecasting of non stationary time series in nonlinear systems, which has some uncertainties, the method has the excellent identification performance and forecasting accuracy superior to other existing linear combining forecasts for the same event.