针对 MIMO复杂过程提出一种通过实验数据获取模糊系统模型的方法 ,即将每一维输入变量的论域进行等间隔分割后确定出模糊规则的前件参数和规则总数 ,再由一种调整算法通过对实验数据的学习得到模糊规则的后件参数。理论分析说明这种模糊规则后件参数学习算法是收敛的、所建模糊模型能够以要求的精度逼近已知的实验数据。仿真举例验证了所述模糊建模方法的有效性并说明了所建模糊规则模型有较好的泛化能力 Aiming at the complicated process of MIMO, a new method to get the fuzzy system model through experimental data is proposed. That is to say, the universe of each input variable is divided equally and the total number of the antecedents and rules of the fuzzy rule is determined. Then an adjustment algorithm is adopted The learning of experimental data obtained after the fuzzy rules of the parameters. The theoretical analysis shows that the algorithm of parameter learning of the consequent of fuzzy rules is convergent, and the built fuzzy model can approximate the known experimental data with the required accuracy. The simulation example validates the effectiveness of the fuzzy modeling method and shows that the built fuzzy rule model has better generalization ability