Research on Carbon Monoxide Content Prediction Based on Improved Particle Swarm Optimization SVM

Abstract

In the petroleum processing industry, the carbon monoxide (CO) content of the flue gas emitted by the heating furnace is predicted according to the various gases that will be generated in the on-site environment of the heating furnace. Therefore, a Particle Swarm Optimization Support Vector Machine(PSO-SVM) model is proposed for the prediction of carbon monoxide content. When optimizing the parameters of SVM for particle swarm optimization, it is easy to fall into the problem of local optimization and premature convergence. An improved particle swarm optimization(IPSO) algorithm is proposed: in the optimization process, adding a passive aggregation term to improve the speed formula of the PSO algorithm can make the particles reach the global optimal state. Finally, the carbon monoxide content is predicted by the improved PSO optimization support vector machine. The experimental results show that the improved particle swarm algorithm is used to fully explore the potential of the SVM model, and compared with the experimental results of other prediction models, it is proved that the model has the advantages of high accuracy.