基于遗传算法的发动机控制通道数学模型辨识

Mathematic model identification of aero-engine main control passage based on genetic algorithm

由于控制器结构的变化,某型发动机电子调节器主要控制通道的动态响应过程为分段线性数学模型。为了获得精确的数学模型,对基本遗传算法进行了改进,提出了一种基于改进遗传算法的辨识算法,能有效辨识出系统的结构跳变时刻及其分段数学模型参数,仿真结果表明,该算法辨识的模型与实际系统接近,较之其它辨识算法具有计算稳定、辨识精度高和适应性强等优点。

Due to the change of controller structure, the dynamic response of main control passage for aero-engine EEC can be described as two liner mathematic modes. An improved genetic algorithm for system identification was developed and applied to this variable system. The resuls imply that this method can identify the system structure jumping time and parameters of the two models, and it is close to the actual system. Compared with other identification algorithm, it has higher precision and fine stability as well as adaptability.