The work aims to optimize and improve the loading efficiency of air materials in the work through the analysis on the current air material loading problems, so as to provide ideas and methods for the intelligent loading of army air materials. The initial idealized conditions and the constraints required to be satisfied in the loading process were determined. With the placement order and rotation mode of air materials as genetic values, the number of loading boxes and the deviation of the center of gravity were used as fitness functions to evaluate the solutions. Elite selection strategy was adopted to guide the direction of evolution. The generation and selection of progeny were relatively partial random key crossover and sequence variation, which avoided the conflict of gene value possibly occurring in the process of selection variation. The least square method was used to fit the input parameters of the algorithm of storage, transportation and loading of air materials, and the optimal input values of the balance time and the number of containers used were found. By comparing with the actual manual loading, the average utilization rate of the loading box was increased from 78.1% to 84.45%, and the classification of air materials was completed. This method has the ability to optimize the loading scheme of air materials, can improve the loading efficiency, and is of great significance for the development of intelligent loading of army air materials.