Adaptive Control Of Aircraft Engine Based On Heuristic Optimization Algorithm

Aircraft engine is a very complex and nonlinear controlled object with strong uncertainand time-varying, so it is difficult to establish its precise mathematical model. Applying thetraditional control method to aircraft engine control is generally difficult to obtain satisfactoryresults. Therefore, those control algorithms which do not depend on the accuratemathematical model of the controlled object have become an important research direction.Firstly, we designed the control method which combines genetic algorithm with PIDcontroller. According to the experience, the approximate scope of PID parameters is estimatedpreliminarily, and then the optimal PID parameters are gotten by using genetic algorithm tosearch optimization in the approximate scope. In order to meet the requirements ofperformance indicators, we designed the aircraft engine control system by combining geneticalgorithm with PID controller.Then, a method of neural network adaptive control was designed based on geneticalgorithm (GA). genetic algorithm was applied to determine initial weights of neuralnetworks(NNs), and NNs was applied to identify the model of engine real-time. Moreover, thesensitivity information of the system was fed back to the NNs controller, which ensured theaccurate control for the engine. With an aircraft turbojet engine, the control system wasdesigned on the selected design point. The simulation results showed that the control systemnot only had good good response speed and overshoot, but also had strong anti-interferenceability.Finally, we designed a NNs control system based on backstepping. Because of thequality of air and fuel of engine is measured, so NNs observer is designed, and estimated thestates variable of the engine by the observer. Then, we designed NNs controller based onbackstepping. The simulation results showed that the control system effectively reduce theengine combustion cyclic variations.