模糊控制是人工智能的一个重要研究领域，已在家电和工业控制中得到重视和应用。模糊规则自动生成是模糊控制的关键性技术之一。模糊控制的目标就是使得一系列输入与期望输出之间的平滑过渡。对于人工神经网可应用于模糊规则自动生成，这是因为人工神经网可用作近似化工具实现这一系列输入与期望输出之间的数据拟合。提出基于多层前向网络的模糊规则自动生成算法（通过数据拟合、有监督的分类），并说明其特点。 Fuzzy control is an important research area of ​​artificial intelligence, which has been valued and applied in home appliances and industrial control. Automatic generation of fuzzy rules is one of the key technologies of fuzzy control. The goal of fuzzy control is to make a smooth transition between a series of inputs and the desired output. Artificial neural networks can be applied to the automatic generation of fuzzy rules because artificial neural networks can be used as an approximation tool to achieve data fitting between this series of inputs and expected outputs. An automatic fuzzy rule generation algorithm based on multi-layer forward network (through data fitting and supervised classification) is proposed and its characteristics are described.