强调模糊性和疑惑性的Atanassov直觉模糊集在决策、图像处理、聚类和专家系统等智能系统中得以广泛的应用。为了更好地实现直觉模糊推理,研究直觉模糊推理系统诸如鲁棒性的基本性质意义非凡。本文给出了直觉模糊t-模和s-模、否和几种蕴涵(即R-、S-和QL-蕴涵)的灵敏度表达式。并根据这些模糊连接词和直觉模糊集的灵敏度深入分析了直觉模糊推理系统的鲁棒性,并发现直觉模糊推理系统的鲁棒性取决于其构成的直觉模糊连接词。 Atanassov intuitionistic fuzzy sets that emphasize fuzziness and confusion are widely used in intelligent systems such as decision making, image processing, clustering and expert systems. In order to achieve intuitionistic fuzzy reasoning better, it is significant to study the basic properties of intuitionistic fuzzy inference system such as robustness. In this paper, the sensitivity expressions of intuitionistic fuzzy t-modules and s-modules and several implication (ie R-, S- and QL-implication) are given. According to the sensitivity of these fuzzy connectives and intuitionistic fuzzy sets, the robustness of intuitionistic fuzzy reasoning system is deeply analyzed. It is found that the robustness of intuitionistic fuzzy reasoning system depends on the intuitionistic fuzzy connectives formed by them.