橡胶隔振器是航空发动机辅助系统的重要连接与减振部件，其隔振性能易受材料老化及服役环境不确定性影响，亟需开展可靠性评估。然而，传统依赖大量随机样本的蒙特卡洛方法计算代价高昂，面对隔振器此类结构复杂、不确定因素多且试验代价高昂的典型航空部件，现有方法仍难以实现兼具计算效率与统计精度的可靠性评估。为此，提出一种基于矩积分与最大熵理论的不确定性量化方法，实现橡胶隔振器隔振性能的高效可靠性评估。首先，依据橡胶材料加速老化试验数据建立弹性模量退化模型，并通过矩积分构建 Hankel 矩阵求得关键采样点与权重，以少量关键样本代替大规模随机抽样。随后，将关键采样点对应的弹性模量输入随机振动有限元模型，获取隔振率响应及其统计矩，并利用最大熵原理在无需预设分布形式的条件下重构隔振率概率密度函数。本方法具有非采样、解析求积的显著优势，仅需少量最优关键采样点即可获得隔振性能的多阶统计特征，在保持精度的同时极大降低有限元调用次数，为复杂结构可靠性分析提供了一种高效的技术路径。

Rubber isolators are important connecting and vibration-damping components in the auxiliary systems of aero-engines. Their vibration isolation performance is susceptible to material aging and uncertainties in service environments， making reliability assessment an urgent need. However， the traditional Monte Carlo method， which relies on a large number of random samples， incurs high computational costs. For typical aviation components like isolators-characterized by complex structures， multiple uncertain factors， and high test costs-existing methods still struggle to achieve reliability assessment with both computational efficiency and statistical accuracy. To address this issue， The study proposes an uncertainty quantification method based on moment quadrature and maximum entropy theory， enabling efficient reliability assessment of the vibration isolation performance of rubber isolators. First， an elastic modulus degradation model is established based on the accelerated aging test data of rubber materials. Optimal integration nodes and weights are derived by constructing a Hankel matrix through moment quadrature， replacing large-scale random sampling with a small number of key samples. Subsequently， the elastic moduli corresponding to the integration nodes are input into the random vibration finite element model to obtain the vibration isolation rate response and its statistical moments. The maximum entropy principle is then used to reconstruct the probability density function of the vibration isolation rate without the need to preset a distribution form. This method features significant advantages of being non-sampling and analytical quadrature. It can obtain multi-order statistical characteristics of vibration isolation performance with only a few optimal integration nodes， greatly reducing the number of finite element calls while maintaining accuracy. It provides an efficient technical approach for reliability analysis of complex structures.