论文部分内容阅读
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 .
Based on the fact that fuzzy reasoning system can approximate any nonlinear continuous function in compact set, a new method of nonlinear combination modeling and prediction based on Takagi sugeno fuzzy rule base is proposed to overcome the shortcoming of linear combination forecasting method in solving non-equilibrium time Sequence combinatorial modeling problems encountered in the difficulties and shortcomings, and gives the corresponding optimization algorithm based on learning automata hierarchy to determine the fuzzy system parameters and fuzzy subset of the division, the theoretical analysis and a large number of economic forecasting examples It shows that this method has strong learning and generalization ability, and it has a good application in dealing with the combination modeling and prediction of nonlinear systems with some degree of uncertainty such as economic time series.