混合遗传算法在气动力参数辨识中的应用

Application of Hybrid Genetic Algorithms for Aerodynamic Parameter Estimation

遗传算法是求解复杂系统优化问题的一种有效方法,具有较强的鲁棒性和全局寻优能力,但计算量大,效率较低。将遗传算法与一维局部寻优算法相结合,构造了一混合遗传算法,并将其用于气动力参数辨识,以取代通常采用的梯度类优化算法。采用该混合遗传算法对某型飞机的横向气动力参数进行辨识计算与分析,结果表明该混合遗传算法是气动力参数辨识的一种有效方法,与遗传模拟退火算法相比,其计算效率有较大提高。

Genetic Algorithms(GA) is an effective method to solve the optimization problems of complex systems owing to its robustness and ability to find the globally optimal point. But the computational cost of GA is great and the efficiency is limited. So a hybrid GA, which combines the GA with 1-dimensional local optimization algorithm, is constructed in this paper and applied in aerodynamic parameter estimation to replace those conventional gradient-based optimization methods. After implementing this hybrid GA to estimate seven lateral aerodynamic parameters of an airplane, it is shown that this hybrid GA is an effective method to estimate aerodynamic parameters, and it is much more efficient than the formerly-used genetic simulated annealing algorithm, combination of the GA and the simulated annealing algorithm.