A Mixed Gas Composition Identification Method Based on Sample Augmentation

In the mining process of coal, oil and natural gas, there are often a large number of toxic, hazardous and explosive gases, and these gases usually exist in the form of mixtures. When these flammable gases reach a certain concentration in the atmosphere, they will explode when they encounter an open flame. Therefore, the accurate detection of mixed gases is of great significance to the prevention of safety accidents. In order to improve the accuracy of mixed gas component recognition, this paper proposes a mixed gas composition identification method. This method uses kernel principal component analysis (KPCA) to extract the characteristics of the responses of gas sensor array. Combining cost sensitive learning (CSL) and synthetic minority oversampling technique (SMOTE) to artificially expand the number of unbalanced samples can effectively improve the classification performance of SOFTMAX. The experimental results show that the proposed method has a higher recognition rate after expanding the samples than before. The proposed method has an average recognition rate of 96.3% for mixed gas components.