| The rapid development of modern technology makes the aircraft engine technology constantly improving. Engine state variable model is established based on the component-level model and is used to research the coupling relations inside the engine.The small perturbation method is utilized to calculate the initial value of the linear model. Based on the model built by small perturbation method, the engine multi-input multi-output linear model is obtained by fitting the component-level model. We employed the appropriate weight to improve the precision of the linear state variable model.We used the diagonal matrix decoupling method to decouple the double-input double-output system. The relative gain matrix is used to show the strength of coupling between the variables. Considering the coupling and the time-delay, we designed a decoupling PID controller. By the interaction index, we investigated the optimization of input-output matching. Based on the new matching, we designed a guaranteed cost controller. Considering the coupling between the variables of the aero-engine, we explored the decoupling neural network control design. A decoupled reference model, the command, the output and the control of the engine were used to train the neural network control online. We also discussed the affection the training step on the system performance and optimized the training step.The computer simulation for the two closed loop system above is carried. Its results show the controlled system is stable and has the satisfying dynamical performance and still performance. |
