In the study of mixed gas identification, a gas classification recognition algorithm based on convolutional neural network is proposed to solve the problem that it is difficult to achieve ideal accuracy when electronic nose is used in the monitoring of chemical pollutant types. Firstly, the adaptive feature extraction ability of convolutional neural network is used to effectively reduce the impact of original data on subsequent operations. Secondly, several experiments are conducted to optimize the parameters of the convolutional neural network to improve the performance of the network model. Finally, the proposed convolutional neural network algorithm and BP neural network algorithm are applied to the experimental data of carbon monoxide and ethylene mixture gas in the public dataset of the University of California, respectively. The experimental results show that the gas species detection accuracy of the convolutional neural network algorithm in this dataset reaches 93%,which is higher accuracy and smaller error than the BP neural network algorithm when applied to gas identification, which provides a new method for gas species detection in the electronic nose system.