提出一种直接利用均匀分布于待逼近系统输入空间的 I/ O数据 ,快速构造满足一定精度要求的模糊逻辑系统方法 ,并从理论上证明了该方法的可行性。在此基础上采用一种新型的 GA+BP混合算法对模糊逻辑系统进行优化 ,以求用最少的规则数实现满意的精度。数字仿真结果表明这种快速构造和优化方法是可行和高效的 This paper presents a method of directly using I / O data which is uniformly distributed in the input space of the system to be approximated and quickly building a fuzzy logic system that meets certain accuracy requirements. The feasibility of this method is proved theoretically. On this basis, a new GA + BP hybrid algorithm is used to optimize the fuzzy logic system in order to achieve satisfactory accuracy with the minimum number of rules. The numerical simulation results show that this rapid construction and optimization method is feasible and efficient