Robust Fuzzy Control For Aeroengines

With the increasing improvement of modern advanced aircraft engines,control systems of aircraft engines become more and more sophisticated.Conventional fuzzy approaches have attracted much attention on aircraft engines and made great achievements,but the control parameters are depended on a series of simulations and experiments.The control based on the fuzzy model can avoid this problem effectively.Therefore,we develop a fuzzy model and propose a robust fuzzy control for aircraft engines.For the strong nonlinearity of aircraft engines,the state space model?SSMs?is extracted by the perturbations and fitting methods from a component-level nonlinear engine model.The uncertainty,including nonlinear dynamic characteristics,noise and disturbance,is described by fuzzy sets.A nominal robust control and a deterministic compensator are designed for the uncertain fuzzy dynamic system,which guarantee that the controlled system is uniformly bounded and ultimately uniformly bounded.The control gain is optimized under a synthetic index.The simulation results show desired performance of the controlled system.Considering that the characteristics of engines vary with the flight conditions,a multivariable modeling method based on the T-S fuzzy model is proposed.The flight conditions are chosen as antecedent variables,and engine SSMs are concluded variables.For the same flight condition,the variation of operating characteristics is considered as the uncertainty.Therefore,an uncertain multivariable T-S fuzzy model for aircraft engines is formed.Further,a non-quadratic control scheme consisting of aH₂/H_?control and a Leitamnn compensator is proposed.The uniform boundedness and the ultimate uniform boundedness,which are more relaxed and practical than an asymptotic stability,are achieved.Considering the adjusting plan of throttle operating condition of a turbofan,a multivariablecontrol frame for the full envelope is constructed based on the Labview software.The hardware-in-loop?HIL?simulations and the semi-physical simulations of the proposed control are performed.The simulation results show the practicability and desired performance of the new robust control methods.The reasons that result in the difference among the three kinds of simulations are analyzed and it would be helpful for the further improvement of control design.