Concentration Prediction of Multi-component Gases Based on Improved Sparrow Search Algorithm

Abstract

The electronic nose consists of a sensor array and software algorithms, which can be used for gas identification and concentration prediction. The least-squares support vector machine (LSSVM) is often used for gas concentration prediction. However, the effectiveness of its performance is largely influenced by the chosen parameters. Hence, it is essential to integrate efficient optimization algorithms to enhance the predictive capabilities of LSSVM. This work introduces an improved sparrow search algorithm (ISSA) to enhance the precision of gas concentration prediction by LSSVM. The ISSA approach is evaluated against the particle swarm optimization algorithm (PSO) and the sparrow search algorithm (SSA). Response data of different concentrations of CH4, CO, H2 and their binary gases mixture have been measured using an electronic nose composed of six gas sensors. After processing feature extraction and principal component analysis (PCA) on the original data, it is used as a training and test dataset for prediction models. The results demonstrate that ISSA can significantly enhance the precision of the LSSVM model for gas concentration prediction.