A De-noising and Compress Algorithm of Multi-sensor System Based on BP Neural Network

摘要

White noise is unavoidable in sensor's output signal. Because of its nonlinear mapping ability, BP neural network can be applied to restrain noise in multi-sensor inspection system. Sensors' output signal and target value can be considered as training sample. Through training and simulating, by means of MATLAB and compared with optimal weight filter method, BP neural network provides better accuracy in restraining white noise. Also, output signal from multi-sensor can be compressed into a fusion result.