GAS RECOGNITION METHOD BASED ON COMPRESSIVE PERCEPTION THEORY

摘要

A gas recognition method based on a compressive perceiving theory, the method comprising: collecting compressed data in an under-sampling manner; reconstructing the collected compressed data to obtain reconstructed data; using the reconstructed data to train a back-propagation neural network and saving the trained back-propagation neural network; inputting data to be detected into the trained back-propagation neural network, and the trained back-propagation neural network recognizing the data to be detected to realize qualitative recognition of the gas. The method solves the problems of large volumes of transmission storage data and imprecise recognition in current gas detection, and achieves the target of using a small volume of data to realize precise qualitative recognition.