Design Of Intelligent Controller For Variable Cycle Engine

The Variable Cycle Engine(VCE)has high unit thrust and low fuel consumption,and can meet the characteristics of the aircraft’s different flight power requirements.It is the first choice for future aircraft power plants.Compared with the traditional aero engine,the variable cycle engine system has a complex structure and more adjustable components.The traditional multi-variable control law of the aero engine is difficult to meet the control requirements of the variable cycle engine.This paper takes a certain type of variable cycle engine as the research object,respectively uses the brain emotion learning algorithm and the improved deterministic strategy gradient algorithm to design the multi-variable control law of the variable cycle engine,and performs simulation verification based on the variable cycle engine model.The main research content include:This paper proposes a variable-cycle engine multi-variable control algorithm based on the brain emotion learning model of the differential evolution algorithm.First,the variable-cycle engine two-variable and three-variable control architectures based on the brain emotion learning model are designed respectively.Secondly,for the problem of brain emotion learning algorithm with the increase of control variables and the sharp increase of controller parameters,a brain emotion learning algorithm based on differential evolution algorithm is proposed,which realizes variable cycle engine multi-component combination adjustment.The simulation results show that the brain emotion learning algorithm based on the differential optimization algorithm can achieve fast and precise control of variable multi-variable engine.This paper proposes a variable-cycle engine multivariable control algorithm based on the deterministic policy gradient algorithm(DPG algorithm)of prioritized experience replay.The proposed method can improve the control precision of the algorithm and speed up the training speed of the control strategy,and realize the non-decoupling control of the eight control variables of the variable cycle engine.Simulation results show that the DPG algorithm based on prioritized experience replay can achieve variable-cycle engine steady-state control.Compared with the traditional multi-variable open-loop or decoupling control method,this method does not require decoupling control and prior knowledge guidance.It is more practical.