基于双模型滤波算法的环控系统换热器故障诊断

 在建立飞机环控系统数学模型的基础上，提出采用双模型滤波方法进行参数估计、状态预测和故障诊断，提高飞机环控系统故障诊断的快速性和准确性。如果采用最小二乘算法，参数估计是静态的，故障诊断延迟一般较大；采用单模型扩展Kalman滤波算法，虽然能够实现动态估计,但不能同时兼顾稳态过程和过渡过程（突发故障）的参数估计，导致误差较大。为了解决上述难题，针对飞机环控系统换热器故障诊断，提出两模型滤波算法。该算法由两个滤波器组成，分别用于跟踪系统的稳态和过渡过程。由于采用了两滤波器模型分别匹配不同的系统特征，能够改善飞机环控系统不同状态下的参数估计和状态预测性能，从而提高系统故障诊断的精度和速度。仿真结果证实了该算法的有效性。

Fault Diagnosis Algorithm for Heat Exchanger of Aircraft Environment Control System Based on Double Model Filter

A fault diagnosis algorithm based on the double model filter(DMF) for a heat exchanger of the aircraft environment control system(ECS) is developed to improve the capability of fault diagnosis.The minimum meansquare values currently used in fault diagnosis can only get the static estimate values,and although the extended Kalman filter(EKF) can realize dynamic parameter estimation,yet for a nonlinear system,a single EKF does not exhibit good ability for either the normal process or the fault process.For the aircraft ECS,a fault diagnosis algorithm based on DMF is presented in this paper to deal with nonlinear systems.Two filters are used to track the different processes(normal or fault) in the DMF.For the ECS,two filters are used respectively to trace the steady state and the transient state of the system.The performance of parameter estimation and state prediction for ECS can be improved by using this method,and the accuracy and speed of fault diagnosis are also improved.Simulation is done to demonstrate the performance of this fault diagnosis algorithm.