Improved hemispherical resonator gyroscope signal denoising method

Abstract

The invention provides an improved hemispherical resonator gyroscope signal denoising method. The method comprises the following steps: firstly, decomposing a measurement signal of a hemispherical resonator gyroscope into a series of intrinsic mode function components by utilizing a self-adaptive noise complete set empirical mode decomposition method; and calculating the variation of the Shannon entropy of adjacent intrinsic mode function components by using the Shannon entropy, finding an intrinsic mode function value corresponding to the maximum entropy variation, separating effective signal components from noise signal components, and finally completing signal reconstruction. The method belongs to a signal processing method in the technical field of automation, and can improve the measurement precision and sensitivity of a hemispherical resonator gyroscope on the premise of not changing the precision of an inertial device of the hemispherical resonator gyroscope.

Description

technical field [0001] What the present invention designs is a hemispherical resonator gyro (Hemispherical resonator gyro, HRG) signal denoising method, more precisely, is to use Shannon (Shannon) entropy to realize adaptive noise complete set empirical mode decomposition (Complete Ensemble Empirical Mode Decomposition withAdaptive Noise, CEEMDAN) denoising method improvement, and use the improved CEEMDAN denoising method to preprocess the output signal of the hemispherical resonant gyroscope, which belongs to a signal processing method in the field of automation technology, and realizes it without changing the precision of the inertial device. Improving the measurement accuracy and robustness of a hemispherical resonator gyroscope. Background technique [0002] The hemispherical resonant gyroscope is a new type of inertial navigation level solid gyroscope. It is a new type of vibrating gyroscope that uses the Coriolis Effect (Coriolis Effect) of the standing wave excited on...

AI Technical Summary

Problems solved by technology

The premise of the optimal estimation of Kalman filtering is that the system model is known accurately, but the noise of the hemispherical resonator gyro in the actual system has strong uncertainty, and the system cannot be accurately modeled, thus affecting the denoising effect; denoising based on wavelet analysis The method has achieved certain results in the field of signal denoising by virtue of its multi-resolution characteristics, but the denoising effect is completely dependent on the selection of wavelet basis functions, so this method has certain limitations in practical applications; the EMD method is An adaptive signal processing method, which relies on the time scale of the data itself to decompose the signal multiple times adaptively, and does not need to pre-set the basis function during the decomposition process, which has great advantages in processing dynamic signals, but the method There are disadvantages of prone to modal aliasing

In recent years, in order to further improve the denoising performance of the EMD method, scholars such as Wu proposed an integrated empirical mode decomposition (Ensemble Empirical Mode Decomposition, EEMD) method to eliminate the modal aliasing problem of the EMD method by adding auxiliary noise. However, the EEMD method still has the disadvantages of not completely eliminating the white noise and increasing the error of the reconstructed signal in the signal decomposition process; Yeh et al. proposed a complementary integrated empirical mode decomposition by adding auxiliary noise in the form of positive and negative pairs. (Complementary EEMD, CEEMD) method to eliminate the residual auxiliary noise in the reconstructed signal and reduce the influence of mode aliasing, but this method cannot solve the problem that different numbers of modes may be decomposed when adding different noises; for this reason, Torres proposed the CEEMDAN method, which can not only alleviate the modal aliasing phenomenon, but also accurately reconstruct the original signal by adding adaptive white noise.

Images