Abstract
A novel fusion algorithm is proposed based on Improved Adaptive Wavelet Threshold De-noising (IAWTD), C-means Support Vector Machine (CSVM) and Ensemble Empirical Mode Decomposition (EEMD) method to eliminate the humidity drift of MEMS gyroscope. Firstly, the IAWTD method is employed to decrease the humidity drift component in MEMS gyroscope output signal. Then, the humidity drift compensation model is established: the input elements are the relative humidity, the change rate of relative humidity and the humidity drift, and the output is the compensated MEMS gyroscope output signal by EEMD method. In order to verify the compensation effect of the fusion algorithm, the gyroscope outputs are collected and analyzed with the relative humidity ranged from 40% to 90% based on the temperature varying from 20°C to 60°C. The results show that the IAWTD-CSVM-EEMD method significantly reduces the influence of relative humidity drift on the gyroscope output, according to the quantitative analysis of Allan variance, the quantization noise of the gyroscope output decreases by 87.78%, 96.37%, 97.77%, 99.17% and 92.62% respectively under the relative humidity ranging from 40% to 90%, as the temperature rose from 20 °C to 60 °C at intervals of 10 °C. In addition, the bias stability decreases by 96.9%, 99.41%, 99.1%, 99.46%, and 99.78% respectively and the angle random walk decreases by 88.16%, 96.54%, 98.16%, 94.43%, and 92.05% respectively at different temperatures. It is worth mentioning that, to further verify the applicability of the fusion algorithm, a group of comparative experiments are added to consider the influence of temperature changes on the gyroscope output under different relative humidity. The experimental results show that the quantization noise, bias stability and angle random walk of the MEMS gyroscope are significantly reduced compared with the original output after processing by IAWTD-CSVM-EEMD. Therefore, the method proposed in this paper is beneficial to reduce the humidity drift in the MEMS gyroscope output.
Highlights
With the development of the MEMS technology, inertial devices including accelerometers and gyroscopes are more and more widely used in inertial navigation
The threshold function of wavelet threshold denoising is improved to get the improved adaptive wavelet threshold denoising (IAWTD) method, and the IAWTDEEMD-C-means Support Vector Machine (CSVM) method is obtained by using C-means to improve support vector machine and combining with Ensemble Empirical Mode Decomposition (EEMD) denoising algorithm
(2) The C-Support vector machine (SVM) prediction model is trained in advance, relative humidity, relative humidity change rate and relative humidity drift extracted by IAWTD are taken as the input of the model, and the model output is subtracted from the
Summary
With the development of the MEMS technology, inertial devices including accelerometers and gyroscopes are more and more widely used in inertial navigation. There are other proposed methods to solve the temperature drift of MEMS gyroscopes such as Radial basis function neural network (RBF-NN), Back propagation neural network (BP-NN), Kalman filter, Wavelet threshold denoising, Support vector machine (SVM) [4,5,6,7,8,9,10]. These studies mainly focused on how to reduce the temperature drift of MEMS gyroscopes, the detailed influence of humidity on MEMS gyroscopes is still lacking. The Allan variance analysis method was utilized to compare the performance of the MEMS gyroscope and improve practicability and significance of the suggested methods
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
