PSO-SVM-based method for predicting the demand for airline material carriage

As an important military aircraft material, with the increase in the number of off-site flight missions, it is inevitable to transfer to other sites for carrying security. With the increase of joint exercises, target training and other missions, the mission environment is complex and changeable, which puts forward higher requirements on the level of carry guarantee, among which the accurate prediction of the demand for aviation materials in the mission is one of the main elements of good carry guarantee. In this paper, we propose to use Support Vector Machine (SVM) to predict the number of aircraft materials to be carried, and apply the Particle Swarm Algorithm (PSO) to optimize the SVM parameters. By simulating and comparing PSO-SVM, GS-SVM and GA-SVM, the PSO-SVM algorithm is able to predict the airline material carrying requirements more accurately in a shorter period of time.