Selection of Wavelet Decomposition Level in Multi-Scale Sensor Data Fusion of MEMS Gyroscope

Abstract Because the characteristic of wavelet transform is multi-resolution, their unique advantage of the data model is multi-scale analysis. Then, it’s widely used the Wavelet-based multi-scale sensor data fusion technology in many fields. There are two problems in operating process, which are the choice of wavelet base and the choice of decomposition level. The wavelet decomposition level is very important to the analysis and will influence the result. Based on the high-frequency detail coefficients of wavelet decomposition and it’s wavelet variance, an improved adaptive determination method to choose the optimal decomposition level is proved in the paper, and the fusion result was effetely show the advantage of the program.