Multi-Component Gas Analyzing Based on Support
Vector Machine
Lin Jipeng,Liu Junhua
(School of Electrical Engineering,Xi'an Jiaotong University,Xi'an 710049,China)
Abstract:According to the regularization theory and the optimal quadratic programming
algorithm,a quantitive analysis method based on least square support vector machine is
proposed to solve the problem of multi-gas analysis,which is mainly restricted by lack of
experimental samples.The perfect nonlinear mapping ability ensures zero training error and
leads to the global optimal parameters, hence the cross-sensitivity of gas sensor is
preferably eliminated among multi-gas components.The multi-gas analysis shows that the
cross-sensitivity is decreased to 1/81 of non-processed.
Keywords:support vector machine;statistical learning; infrared analyzing apparatus;gas
analyzing