遗传算法在控制率参数辨识中的应用

Application of genetic algorithms in control law parameters identification

为解决现阶段飞机控制律设计和调整很大程度上依赖于设计人员自身经验的问题,提出了一种应用遗传算法辨识控制率参数的方法。首先以国产某型号飞机为研究对象设计爬升控制律,然后应用遗传算法,对控制率中的未知参数构造合适的解码公式并且设定相应的适应度函数,最后基于机载先验飞行数据进行运算,通过判定适应度函数值,得到优化后的未知参数。仿真结果表明,飞机爬升至巡航后,俯仰角与空速稳定,巡航高度偏离目标高度不超过1.5%,这种设计方法得到了比较高的控制精度,能够满足实际应用的需要。

According to the problem that the design and adjustment of aircraft control law relies on experience to a great extent at present, applying with genetic algorithms, one method of parameter identification is put forward. First, design climb control law of a certain domestic aircraft. Then, construct suitable decoding formulas for unknown parameters of control law and set corresponding fitness function. Finally ,basing on airborne prior flight data, by judging fitness function, calculate optimal unknown parameters. The imitation result shows that pitch angle and airspeed are stable, deviation value of cruising altitude is lower than 1.5%.The design method can get good control precision and meet practical needs.