涡扇发动机早期退化性能的线性变参数估计

Linear parameter varying estimation for early deteriorated turbo fan engine performance parameter

针对涡扇发动机线性变参数(Linear parameter varying,LPV)模型,给出了一种新的早期退化航空发动机部件性能LPV滤波器设计方法.该方法采用仿射依赖于参数的二次Lyapunov函数,在仿射二次稳定性的基础上,引入多凸性引理将滤波器求解问题转化为线性矩阵不等式(Linear matrix inequalities,LMI)约束下的凸优化问题,避免了求解参数化线性矩阵不等式(Parameterized linear matrix inequalities,PLMI)约束下的优化问题.针对某型早期退化涡扇发动机的仿真表明,该方法能以较高精度估计性能退化参数,且鲁棒性强,增益调度简单,具有一定的工程意义.

To estimate the component performance parameters of early deteriorated areoengine, a linear parameter varying（LPV）filter was given for LPV model of a turbofan engine. Based on conceptions of affined parameter-dependent quadratic Lyapunov function and affined quadratic stability, the filter design method was converted into a normal linear matrix inequalities（LMI） constrained convex optimization by introducing the lemma of multiconvexity. Also the difficulties in solving the optimization problem with parameterized linear matrix inequalities（PLMI） constraint were avoided. Besides the convenience of filter gain scheduling, the accuracy and robustness of the filter were validated by simulation on parameter estimation for the early-deteriorated turbofan engine.