Study on Improved Singular Value Decomposition De-noising Method Applied to UAV Flight Parameter Data
Yuqian Wang, Heng Chen, Dawei He, Jingbo Peng, Baoxi Yuan, Yang Li · 2019
To solve the noise problem of unmanned aerial vehicle (UAV) flight parameter data, the paper proposed an effective singular value decomposition method optimized by chaos ratio bat algorithm (CRBA). The basic bat algorithm is improved by using sigmoidal map, energy of singular values is taken as the objective function in proposed algorithm to optimize the structure of Hankel matrix. The number of effective singular value determined by using the singular value's singularity detection ability, and the de-noised data obtained from singular value and its corresponding vector. According to the experiment results of linear signal and flight parameter data, the proposed method is applicable to nonlinear signals without obvious characteristic frequency and linear signal, which has good denoising results.