为了探究进气畸变条件下轴流压气机的有效失速预警方法并将其应用于主动控制系统中，在一台低速单转子轴流压气机上对均匀进气和周向畸变条件下的失速预警方法进行了实验研究。实验通过在进口布置高度可调节的插板来产生不同强度的周向畸变，并通过在转子叶顶布置的动态压力传感器测得非定常压力信号，运用自相关、互相关和均方根算法对其进行分析，比较了周向畸变下三种失速预警方法的可靠性。分析结果表明：只有传感器安装在畸变区时自相关和均方根分析才有效；互相关分析不依赖于传感器的安装位置，在畸变区和非畸变区均能有效感受到畸变的影响，且互相关系数随着周向畸变的产生或消失相应地下降或上升，因此在周向畸变下互相关分析为更有效的失速预警方法；最后在基于互相关分析的基础上对该单转子轴流压气机实施了喷气主动控制应用，在24.75%畸变强度条件下，主动喷气能取得9.32%的失速裕度改善，与定常喷气相比，获得几乎相等的失速裕度改善的同时节省了约40%的喷气量。

To investigate the feasible stall warning method for active control application in axial compressors with inlet distortion, stall warning methods were experimentally studied in a low-speed isolated-rotor axial compressor under uniform inflow and circumferential distortion. A flat-baffle was installed with different height to generate different distortion intensity. Unsteady pressures were obtained by high-response transducers at tip region and analyzed by auto-, cross- and root mean square (RMS) algorithms to compare the reliability of three stall warning methods with circumferential distortion. The analytical results show that compared to auto-correlation and root mean square analysis which are reliable only at distorted region, cross-correlation analysis can sensibly react to the circumferential distortion not dependent on the transducers’ location. Moreover, cross-correlation coefficient correspondingly decreases or increases with the generation or disappearance of circumferential distortion. Therefore, the cross-correlation analysis is the most effective one among these three stall warning methods. Additionally, the active control was implemented based on cross-correlation analysis in studied isolated-rotor axial compressor. Active injection can obtain 9.32% stall margin improvement under 24.75% distortion intensity condition, saving about 40% injected energy with nearly an equal stall margin improvement compared to steady injection.

V231.3

国家自然科学基金面上项目（51676184）；国家自然科学基金重大科研仪器研制项目（51727810）；国家自然科学基金重点项目（51636001）；国家重大科技专项（2017-Ⅱ-0005-0018，2017-Ⅱ-0005-0017）。