Improvement of support vector machines for mixed VOCs gas recognition based on an arithmetic optimization algorithm

Metal oxide semiconductor (MOS) gas sensors are widely used in gas sensing due to their high sensitivity, low cost, and excellent stability. However, their cross-sensitivity to similar gases limits their detection accuracy in mixed gas environments. To enhance the qualitative recognition performance of gas sensors, this study focuses on a WO₃-based sensor. An arithmetic optimization algorithm (AOA) is employed to optimize the parameters of the support vector machine (SVM), aiming to improve both accuracy and computational efficiency. Experimental results indicate that the WO₃ sensor exhibits high selectivity toward triethylamine (TEA). When combined with the optimized SVM model, the recognition accuracy for binary gas mixtures exceeds 85%, representing a 6%improvement over the original SVM. This method demonstrates superior classification accuracy and efficiency under complex gas conditions, offering a promising approach for real-time gas detection.