Electronic noses are one of the predominant technological choices for gas mixture detection, but their application in real-world atmospheric environments still leaves several issues to be resolved. The key bottleneck is the effect of turbulence caused by the diffusion of gases in the atmosphere on the quantitative and qualitative analytical performance of the electronic nose. In light of this, this paper presents a quantitative and qualitative analysis strategy for gas mixture monitoring. This strategy adopts baseline manipulation of the raw sensor data to reduce drift interference, and then performs feature extraction on the multidimensional response signals of the MOS gas sensor array using principal component analysis (PCA). In order to improve gas mixture recognition accuracy, the whale optimization algorithm (WOA) is used to optimize the network structure of the long short-term memory (LSTM) model for turbulent gas mixture composition recognition. The least squares support vector machine (LSSVM) algorithm is adopted to implement turbulent gas mixture concentration prediction. This paper focuses on two aspects of hyper-parameter optimization for the development of an LSSVM with particle swarm optimization (PSO) and for improved training sample selection for the LSSVM which should subsequently increase the accuracy of concentration estimation. The effectiveness of the proposed strategy is evaluated with a dataset from a chemical sensor array exposed to turbulent gas mixtures. Experimental results revealed that the proposed strategy for turbulent gas mixtures has satisfactory outcomes for both qualitative gas composition recognition and quantitative gas concentration prediction.

Component identification and concentration estimation of a gas mixture component are important for gas detection. However, the accuracy of traditional gas identification will decrease if the sample is not balanced or the number of samples is too few. In this paper, a method based on sample expansion is proposed to solve the aforementioned problem. Firstly, the ADASYN-ELM method is proposed to identify the composition of a gas mixture component. The KPCA is used to extract the feature of the sensor signal and the ADASYN method is used to expand the samples. The PSO and GA algorithms were used to optimize the parameters of the ELM classification model to complete the qualitative analysis. Secondly, the S-SMOTE-MLSSVR method was put forward to quantitatively estimate. The S-SMOTE method was used to expand the samples, and the regression model MLSSVR was optimized by PSO and GA algorithms to complete the quantitative analysis. The results show that the accuracy rate after sample expansion is generally higher and the MAPE and RMSE are generally lower than before sample expansion, indicating that the sample expansion method has a positive effect on classification and concentration estimation of mixed gases with extremely unbalanced samples and too few samples.