基于混和寻优算法的时域低阶等效系统方法

现代高增稳飞机的阶次高达几十阶,需要将其降阶为具有特定形式的低阶等效系统才能与已有的飞行品质规范相比较,从而评价其飞行品质并改进飞控系统的设计.相对于频域等效系统方法,时域低阶等效系统方法可以在线采集系统的输入及输出数据,充分考虑各种非线性因素的影响,在飞机飞行品质评价时,对等效系统相关参数进行在线估计,具有较强的实用价值.针对时域等效拟配提出了一种混和寻优算法,即模式搜索-模拟退火法,该算法较好地克服了局部寻优和全局寻优的缺点,并在寻优过程中加入了对参数范围的限制,寻优结果可以快速收敛于全局最优解,并且不受初值变化的影响,对飞机飞行品质的在线评价具有重要意义.

Time-domain low order equivalent system base on hybrid optimization algorithm

The order of modern highly augmented aircraft is very high.It should be reduced to it's low order equivalent system(LOES) for evaluating the handling quality of aircraft and the design of flight control system.The time-domain equivalent matching can acquire on-line data of input and output,considering the influence of nonlinearities in the system,and the on-line estimation of equivalent parameters is given,it has more practical value than the frequency-domain method.A hybrid optimization algorithm is introduced for the time-domain equivalent matching,namely direct search-simulated annealing method.This method improves the local and global optimization with the limitations to the range of parameters in the process of optimization and can converge the globally optimal solution quickly.The result is not influenced by the initial value and is significant for the on-line evaluation of aircraft handling qualities.