Abstract: For the purpose of exploring better global optimization algorithm and applying it on the aerodynamic optimization of axial compressors， the exploration ways of employed bees and onlookers bees in standard artificial bee colony （ABC） algorithm were improved to make better use of the overall bee colony exploration information . The improved artificial bee colony （IABC） algorithm was obtained in this way. This algorithm was tested by benchmark functions and the results showed that it can both enhance the ability of global optimization and speed up the convergence rate. The IABC algorithm and the CFD simulation method which was checked were applied on the optimization of single stage transonic axial compressor Stage35. The optimization variables included sweep， lean， Re-camber at the leading edge and the trailing edge at the six sections along the radial direction of the rotor blade and stator blade. The relative change of flow rate and pressure ratio within 0.5% were kept as a constraint and enhancing the adiabatic efficiency was as the optimization goal. The results were as follows: at the designed speed， the optimized adiabatic efficiency at designed point increased by 0.83%， the average adiabatic efficiency over whole range of condition increased by 2.0%， and the surge margin increased by 1.0% under the condition that the mass flow and the pressure ratio basically remained the same， and it verified the effectiveness of the optimization algorithm in axial compressor.

Key words: Improved artificial bee colony algorithm;Global optimization;Convergence speed;Axial flow compressor

摘要： 为了探索更优的全局优化算法并将其应用于轴流压气机气动优化上，将标准人工蜂群（ABC）算法中采蜜蜂和观察蜂探索新蜜源的方式进行改进，从而更好地利用总体蜜源的探索信息，得到改进人工蜂群（IABC）算法。经基准函数测试表明，改进算法既能提升全局寻优能力，又能加快收敛速度。采用IABC算法和经过校核的CFD数值方法搭建优化平台，对单级跨声速轴流压气机Stage35进行优化。优化变量为动叶和静叶径向6个截面的弯、掠、前缘重弯值和尾缘重弯值，以流量和压比相对变化保持在0.5%以内为约束条件，以提高绝热效率为优化目标。结果表明：在设计转速下，优化后设计点绝热效率提升0.83%，全工况范围内平均绝热效率提升2.0%，同时喘振裕度提升1.0%，验证了IABC算法在轴流压气机优化中的有效性。

关键词: 改进人工蜂群算法;全局寻优;收敛速度;轴流压气机