基于T-S模糊模型的航空发动机非线性分布式控制系统故障诊断

针对航空发动机非线性分布式控制系统的故障诊断问题,首先提出了一种基于飞行包线划分的航空发动机非线性Takagi-Sugeno(T-S)建模方法,建立了具有网络诱导时延的航空发动机非线性分布式控制系统模型,然后将该系统视为离散切换系统,为其建立了具有时延补偿功能的故障观测器,给出了使得观测器误差系统渐近稳定的充分条件.故障检测仿真时间为20s,当第12s时,设定系统发生幅值为0.0025的阶跃型突变故障,仿真结果表明:12s之前,故障观测器保持渐近稳定,当第12s时,残差迅速增大并超过所设定阈值,从而检测到故障的发生. 

Fault detection of aero-engine non-linear distributed control system based on T-S fuzzy model

The problem of fault detection for aero-engine no-liner distributed control system was studied.Firstly,a new non-linear modeling method for aero-engine based on Takagi-Sugeno(T-S) fuzzy model and flight envelop division was proposed,and the model of the distributed control system with network derivational time-delay was established by using this method.Secondly,if the distributed control system was regarded as a discrete switch system,a fault detection observer with time-delay compensation was designed,and the sufficient conditions of asymptotically stability for the observer-based residual system were presented.The total simulation took 20 seconds, a step fault with 0.0025 amplitude happened at the 12th second, the simulation results show that: the fault observer was asymptotic stable for preceding 12 seconds, and the residuals rapidly increased and exceeded the threshold value at the 12th second, thereby the fault would be detected.