The development of air transport industry can not be separated from the support of airports. How to improve the operation efficiency of airports to meet the increasing demand of flight takeoff and landing under the limited hardware conditions has attracted more and more attention. In order to further optimize the allocation of airport gates in airport terminal buildings, this paper presents a model of airport gate allocation considering the composite categories of airport gates. The model describes the optimization of airport gate allocation with apron assistance under the constraints of airline type, flight type and minimum parking interval. Based on the accurate algorithm validation of the model, a more efficient improved harmony search algorithm is proposed to solve the model. In order to adapt to the characteristics of the medium and large scale problems, the improved harmony search algorithm is introduced by adding complex constraints, with improvements on encoding and decoding, and the initial solution generation and optimization process with. By using Lingo and Matlab respectively on small-scale and medium-scale examples, the result shows that the proposed improved harmony search algorithm is more effective and has advantages over the accurate algorithm in global search ability and solution efficiency. By solving large-scale problems, the proposed algorithm reduces the total transit distance of passengers and achieves good results under the condition of guaranteeing the utilization efficiency of the airport gates.