Study On Modeling And Control Technology Of Short Takeoff And Vertical Landing Engine

Short take-off and vertical landing(STOVL) aircrafts not only possess the characteristic of fast-speed and large combat radius as traditional fixed wing aircrafts, but also depend less on runway and have high flexibility as rotorcrafts. Thus, this type of aircraft has an extensive attention prospect in military field, since the advent of attache’s great importance to by the aviation superpower. While in China, it is in the situation of late-start and little design experience, so there will be a large amount of work to be done about it. This paper mainly researched modeling and controllers design methods of STOVL engine, and digital simulation verification was carried out.Firstly, the modeling method of STOVL engine was studied. Lift-fan mathematical model and adjustable bypass model were built and added to a turbofan engine model. The root and apex of the fan was modeled segmented. The components characteristics were got by the method of characteris-tics modification, and the static and dynamic balance equations were established, then the STOVL engine mathematic model was gotten. To overcome the shortcomings of partial derivative method and fitting method in model linearization, the particle swarm optimization(PSO) method was adopted to establish the STOVL engine linear model, which lays the foundation for the following design of con-trollers.Secondly, the normal mode control method of STOVL engine was studied. Engine working state was divided into idle, throttle and intermediate states according to the power level angle(PLA). The controller based on inverse neutral network and PI algorithm was designed and applied to engine idle state. The augmented linear quadratic regulator(ALQR) was designed and applied to throttle state. The outputs selection and control structure design were studied at intermediate states, and the H2/H∞ controller was design based on linear matrix inequality(LMI). The three type controllers were veri-fied by digital simulation verification.Lastly, the direct thrust control of STOVL engine in hovering mode was researched. The thrust estimators on the basis of Improved Recursive Reduced Least Squares Support Vector Regression(IRR-LSSVR) were designed to map the lift fan thrust and cruising engine thrust respectively. The controller of thrust was designed bused on ALQR method. The efficiency of the control system was tested by simulation.