直觉F推理与普通F推理的比较研究

直觉F推理克服了普通F推理在不确定性信息的描述、推理结果可信性等方面存在的局限性。在介绍普通F推理直觉化扩展的基础上，首先分析了两类推理算法的相互转化问题，指出普通F推理是直觉F推理的一种特例，当直觉指数为0时二者可相互转化。其次，比较了两类算法的还原性，分析表明Zadeh型、Mamdani型、Larsen型直觉F推理算法与其对应的普通F推理算法具有相同的还原性。最后，通过实例研究了直觉F推理算法在推理结果精度、可信性上的优势，从而较普通F推理更适用于智能控制与决策。

Comparative study of intuitionistic fuzzy inference and normal fuzzy inference

Intuitionistic fuzzy inference overcomes the limitation of normal fuzzy inference such as the problem of describing fuzzy information and the creditability of inference results. Based on the introduction of intuitionistic extensions of normal fuzzy inference, the transformation of two inference algorithms were analyzed, which indicated that fuzzy inference was an example of intuitionistic fuzzy inference, and the two methods could transform when intuitionistic index was zero. Then, reductivity analyses showed that Zadeh, Mamdani and Larsen intuitionistic fuzzy inferences had the same reductivity as the corresponding fuzzy inference algorithm. Finally, the performance study by experiment shows that intuitionistic fuzzy inference is more applicable in intelligence control and decision making than normal fuzzy inference.