Abstract:  In view of the fact that the traditional artificial bee colony (ABC) algorithm has a slow convergence speed and traps into local optima easily, a new hybrid artificial bee colony (HABC) algorithm is proposed, which improves the optimization performance in three aspects: selection mechanism, neighborhood search mechanism and diversity of solution vectors. Aimed at the aeroengine control system, an online adaptive PID controller of optimum is designed based on this algorithm. In the control process, the parameters of PID controller are optimized constantly by HABC algorithm so that the aeroengine controller can adaptively obtain the time-varying optimal parameters according to the current system working status. Simulation results show that the aeroengine online adaptive PID controller implements the time-varying optimum of PID controller parameters, which insures that the closed loop system has good dynamic performance as well as strong robustness.

Key words: Hybrid artificial bee colony (HABC) algorithm, aeroengine, online optimization, adaptive, proportional integral derivative (PID) control

摘要： 针对传统人工蜂群算法收敛速度慢、容易陷入局部最优等缺点，提出了一种改进的融合蜂群(HABC)算法，从选择机制、邻域搜索机制和解向量的多样性3个方面改进了蜂群算法的寻优性能。针对航空发动机控制系统，基于HABC算法设计了一种在线自适应PID 控制器，在控制过程中不断优化PID参数，使控制器能够根据发动机系统的当前工作状态自适应地得到时变最优参数。仿真结果表明，所设计的在线自适应PID控制器实现了参数时变最优，使航空发动机闭环系统具有满意的动态性能和鲁棒稳定性。

关键词: 融合蜂群算法, 航空发动机, 在线优化, 自适应, PID控制