基于EKF的涡轮冲压组合发动机跟踪滤波器设计

Tracking Filter Design for Turbine Based Combined Cycle Engine Based on EKF

针对涡轮冲压组合发动机性能参数最优估计问题,介绍了一种基于EKF(扩展卡尔曼滤波)和在线发动机实时模型相结合的跟踪滤波器方案。跟踪滤波器感知模型输出与传感器测量值间的偏差,利用EKF滤波算法动态调整模型状态变量,使模型与发动机间的状态误差方差最小。仿真结果表明,在发动机工作状态发生显著变化时,基于EKF设计的跟踪滤波器能准确跟踪传感器输出,对发动机推力、空气流量、喘振裕度等参数有很高的估计精度。

In order to get the optimal estimation of performance parameters for turbine based combined cy-cle engine （TBCC）, a tracking filter design plant based on extended kalman filter （EKF） and real time on-line engine model was introduced. The tracking filter can dynamically adjust model states according to the errors between on-line engine model outputs and sensor measured values. The variance of state errors can be minimized because of the EKF adjustment. The simulation results show that filter outputs can trace sensor measured values quickly and precisely while engine operate conditions were changed greatly. Mean-while, engine unmeasured parameters, just like engine thrust, mass flow and surge margin, were calculated with high accuracy.