风扇/压气机周向模态识别中传感器角度偏差的影响

基于任意角度压缩感知(CS)方法分析了传感器安装角度偏差对风扇/压气机周向模态识别重构的影响,设计了一套自适应角度优化程序修正重构误差。利用数值试验探究了传感器角度偏差和数量对周向模态重构结果的影响,研究表明:当角度偏差等级为2.5%时,平均重构误差达到10%以上,若保证重构误差基本不变,将传感器数量从7个增加至25个,仅可以将角度偏差等级放宽至4%。而采用小生境微种群遗传算法进行自适应角度优化,在20 dB信噪比下,通过自适应角度优化可将角度偏差等级从2.5%放宽至10%,降低了传感器安装的精度要求。成功优化了一款冷却风扇在前三阶叶片通过频率下的主要周向声模态重构幅值。自适应角度优化算法有效提升了基于压缩感知的风扇/压气机周向模态重构可靠性。

Effect of sensor angle deviation on fan/compressor azimuthal mode recognition,

Based on the arbitrary angle compressive sensing (CS) method, the influence of sensor installation angle deviation on the fan/compressor azimuthal mode recognition was analyzed, and an adaptive angle optimization program was designed to correct the reconstruction error. The influence of angle deviation and number of sensors on the results of azimuthal mode reconstruction was investigated by numerical experiments. When the angle deviation level was 2.5%, the average reconstruction error was more than 10%. If the reconstruction error was basically unchanged, and the number of sensors increased from 7 to 25, the angle deviation level can only be expanded to 4%. The niching micro genetic algorithm was used for adaptive angle optimization. Under 20 dB signal-to-noise ratio, the angle deviation level can be expanded from 2.5% to 10% through adaptive angle optimization, which reduced the accuracy requirements for sensor installation. The reconstructed amplitude of the main azimuthal acoustic mode at the first three-order blade passing frequency of a cooling fan was optimized. It was shown that the adaptive angle optimization method can effectively improve the reliability of fan/compressor azimuthal mode reconstruction based on CS.