Abstract: However, the air contains many components and the content fluctuates greatly, which seriously affects the accuracy of the classification results. In order to increase the reliability of air quality classification, a Particle swarm optimization(PSO) support vector machine(SVM) classification method was proposed. This method firstly searches for the optimal particles globally as the operating parameters of the support vector machine by iterative optimization, and then uses the training set data to perform machine learning to establish a multi-classification model of the support vector machine, and finally imports the input vector of the test set into the model and gets the classification result. The analysis of the results showed that the support vector machine classification method of particle swarm optimization could effectively suppress the influence of artificially set operating parameters on the classification results, improved the classification accuracy of the support vector machine, and provided a new research for air quality classification problem.