基于自适应滑动均值和小波阈值的叶尖间隙信号降噪方法

The tip clearance parameter between the moving blade and the engine casing is one of the key state parameters reflecting the working performance and operational safety of an aero-engine. Improving the signal-to-noise ratio of the tip clearance signal is the key to achieving high-precision tip clearance measurement. For this reason, a real-time hybrid tip clearance signal denoising method based on adaptive moving average and wavelet threshold is proposed. Firstly, according to information such as the blade rotation speed, the bandwidth size of the tip clearance signal is estimated. Then, adaptive low-pass filtering is achieved by dynamically changing the number of moving points of the moving average filter. Finally, combined with the time-frequency analysis ability of wavelet threshold denoising, the noise interference is further reduced. Through simulation experiments, it is determined to use the db5 wavelet basis function for 6-layer wavelet threshold decomposition. The simulation results show that this method is comprehensively superior to the existing finite impulse response and fixed-point moving average filtering methods in various denoising evaluation indicators. The actual tests in the rotation speed range of 1,000 rpm to 4,000 rpm show that the maximum measurement error of this method is 16 μm, effectively improving the measurement accuracy of the tip clearance.

动叶片与发动机机匣之间的叶尖间隙参数是反映航空发动机工作性能和运行安全的关键状态参数之一。提高叶尖间隙信号信噪比是实现高精度叶尖间隙测量的关键,为此提出基于自适应滑动均值和小波阈值的混合叶尖间隙信号实时降噪方法。首先根据叶片转速等信息,估算叶尖间隙信号带宽大小,然后通过动态改变滑动均值滤波的滑动点数来实现自适应低通滤波,最后结合小波阈值降噪时频分析能力,进一步降低噪声干扰。通过仿真实验确定采用db5小波基函数进行6层小波阈值分解,仿真结果表明该方法在各项降噪评价指标方面全面优于现有的有限脉冲响应和固定点滑动均值滤波方法。 1 000 rpm~ 4 000 rpm转速范围内实际测试表明,该方法最大测量误差为16 μm,有效提高叶尖间隙测量精度。