用于控制器分区设计的发动机飞行包线区域最优划分

Partition optimization in the flight envelope for control design of aero-engines

结合最优化理论和线性系统理论,对控制问题中航空发动机飞行包线区域划分方法展开了研究.首先基于小偏离线性动态模型,根据线性系统理论,将发动机动态特性的摄动归为线性模型的状态矩阵特征值和稳态增益的摄动来表征.然后在此基础上,定义了一种广义距离以表征发动机动态特性的摄动程度,并以定义了区域的划分原则.继而提出将飞行包线区域划分问题转化为一个基于该广义距离的最优化问题.最后就某型涡扇发动机情形,对其飞行包线区域进行了划分计算,并对划分区域内发动机线性模型摄动进行了比较,结果表明该区域划分和标称点选择方法的有效性.

According to the optimization theory and linear system theory, the method of flight-envelop optimization partition for aero-engines is researched. First, linear state variable mode （LSVM） was used to describe aero-engine＇s dynamic characteristic of small perturbation on steady operating point, and the perturbation between LSVMs was denoted by the perturbation of coefficient matrix eigenvalues and low-frequency gains of the LSVMs. Then, a general distance based on the perturbation of eigenvalues and low-frequency gains of the LSVMs was put forward. And the problem of dividing the flight envelope and selecting the nominal points in a flight envelope was translated into a most overlay problem. Thus, based on the general distance, a set of least number of nominal points could be found to divide the flight envelope by using optimizing method. Last, a flight envelope area of a turbofan engine was divided by using this method, and the results indicated that this method was proper.