旋转失速限制了压气机稳定工作的范围，对其进行深入理解并实现准确预测是控制失速、提高裕度的关键。现有模型大多基于一定程度的几何和流动简化，不考虑三维效应及流动复杂性，直接应用于三维压气机失速预测时仍面临巨大挑战。同时，尽管试验测量和模拟仿真水平不断提高，试验和数值模拟多为唯象研究，缺乏对压气机流动失稳根本原因的揭示。此外，由于三维复杂流动精细化测量和高保真模拟的复杂性，大多数失速研究针对某一压气机若干孤立工况开展，缺乏系统的参数化研究，难以提炼出旋转失速关键影响因素。为弥补试验测试空间分辨率低和非定常流动模拟成本高的缺陷，提出了一种基于三维流动方程高效特征值求解的全局稳定性分析方法。一方面可以获得试验测量难以达到的空间分辨率，另一方面能够以比非定常模拟小2~3个量级的成本获得丰富的三维流场小扰动发展过程。针对某典型跨声速压气机环形叶栅，所发展的分析方法计算成本仅为定常特性线计算的28%，相比于非定常计算实现了约155倍的加速，为压气机旋转失速准确快速预测和机理研究提供了重要的研究工具。

Rotating stall limits the stable operating range of compressors， and a deep understanding and accurate prediction of this phenomenon is key to stall prediction and control. Existing models for stall prediction are based on simplification of the compressor geometry and flow， and thus applications of these models to stall onset prediction of actual compressors are faced with considerable challenges. Meanwhile， despite the progress in experimental measurements and flow simulations， most experiments and numerical simulations are phenomenological research， and did not reveal the root cause of compressor flow instability. Moreover， due to the complexity of three-dimensional flow measurements and the high cost of unsteady simulations， most stall studies are conducted only under isolated working conditions of a particular compressor， as a systematic parametric study to identify the key influencing factors is too costly. In order to circumvent the shortcomings of both measurements and unsteady simulations， a global stability analysis method based on the efficient eigenvalue solution of the three-dimensional flow governing equation is proposed. This method can obtain not only the spatial resolution that is difficult to achieve by experimental measurement， but also the same rich information of the three-dimensional flow perturbation development at a cost two to three orders smaller than the unsteady simulation. In this paper， for a typical transonic compressor annular cascade， the computational cost of the proposed method is only 28% of that for computing one steady speedline， and 155 times faster than the unsteady calculation. Therefore， the proposed method provides an important research tool for accurate and rapid prediction and mechanism research of rotational stall of compressors.