利用模糊穴映射理论,提出一种有效描述复杂多变量系统的模糊模型——广义模糊基函数展开式。它可方便地处理步输入多输出系统的语言和系统信息,并可逼近任意非线性函数,是一种通用的多变量模糊逻辑系统模型.利用语言信息,提出一种新的自适应参数辨识方法——改进的Widrow—Hogg学习规则,仿真结果验证了它的有效性。 By using the fuzzy-hole mapping theory, a fuzzy model that can effectively describe complex multivariable systems is proposed-generalized fuzzy basis function expansion. It can process verbal and system information of step input multi-output system easily and approximate any non-linear function, which is a general multivariable fuzzy logic system model.Using the language information, a new adaptive parameter identification method - Improved Widrow-Hogg learning rules, the simulation results verify its effectiveness.