摘要
将混合遗传算法应用于飞行器气动参数辨识。该方法结合了遗传算法的全局寻优能力和极大似然法的局部寻优能力,使得混合遗传算法不受极大似然法初值选取的影响,同时也解决了遗传算法收敛速度慢和收敛精度较低的问题。在混合遗传算法寻优过程中,仅对最优个体和变异后的个体执行局部寻优操作,从而使得混合遗传算法的计算量维持在一个适当的水平。最后,一个飞行器纵向模型气动参数的辨识仿真表明:混合遗传算法的收敛性和精度都远高于没有采用局部寻优策略的遗传算法。
Hybrid genetic algorithm is applied to the identification of aircraft＇ s aerodynamic parameters. This algorithm combines the global search ability of genetic algorithm and local search ability of maximum likelihood method, which makes the hybrid genetic algorithm unaffected by the initial value assignment of maximum likelihood method, and solves the problem of the low convergence rate and low convergence precision of genetic algorithm. In the search process of hybrid genetic algorithm, local search operation is performed only to the elitist and the individuals obtained by mutation, thus the hybrid genetic algorithm has an appropriate computation time. At last, an identification example of the aerodynamic parameter of an aircraft＇ s longitudinal model indicates that the convergence rate and convergence precision of the hybrid genetic algorithm are superior to that of the genetic algorithm without local search strategy.
出处
《电光与控制》
北大核心
2006年第2期15-17,23,共4页
Electronics Optics & Control
关键词
混合遗传算法
极大似然法
参数辨识
非线性系统
hybrid genetic algorithm
maximum likelihood method
parameter identification
nonlinear system
