Application of translation-invariant multiwavelets denoising using neighboring coefficients to fault diagnosis of antennas on vessel

In order to solve the problem that antennas on vessel easily arises machinery fault, translation-invariant multiwavelets denoising using neighboring coefficients is researched and applied to diagnosis for failure of satcom antenna. Neighboring coefficient denoising considering relativity of coefficients is introduced to translation-invariant multiwavelets domain to take advantage of excellent properties that scalar wavelet cannot satisfy, avoid Gibbs phenomena and enhance signal denoising effect. Experimental results show that proposed scheme extracts feeble character of early fault more effectively. In practice, damage on the sun gear of azimuth 1# decelerator is successfully diagnosed, which proves effectiveness of translation-invariant multiwavelets denoising using neighboring coefficients to failure diagnosis of antennas on vessel.