Identification And Characterization Of Design Points For Micro Turbojet Engines

Micro turbojet engine is the power device of micro unmanned surveillance aircraft,micro guided munitions and other aircraft.The performance of aircraft is determined by the performance of micro turbojet engine.At present,with the development of the performance of these aircraft,such as strong endurance,high maneuverability,good stealth and miniaturization,higher requirements are put forward for the structure size and thrust-toweight ratio of micro turbojet.The process of engine design point identification is to find the influence law of aero-thermodynamic parameters and size parameters on engine performance index.In order to design micro turbojet engines with small size and large thrust-to-weight ratio,this paper designs the structure of micro turbojet engines,and identifies the design points of micro turbojet engines.Firstly,the aerodynamic design and structural design of the compressor and turbine,which is the core components of the micro-turbojet engine,were carried out.Secondly,the fluid dynamics simulation analysis of the compressor and turbine were carried out.The analysis results used as the sample database for identification of the design point.Then the identification models of centrifugal impeller and radial turbine based on BP neural network were designed.And the neural network toolbox in Matlab to verify the model.The verification results show that the BP neural network identification model designed for centrifugal impeller and radial turbine respectively can identify and characterize the design points of centrifugal impeller and radial turbine.Finally,A micro turbojet engine test bench was designed,impeller blade Angle of incidence with the test bench for corner after 60°,export and air inlet Angle of 45° for centrifugal impeller is 60°,the flow outlet Angle of 30° radial turbine import and export pressure measurement experiment was carried out.The experimental results show that the inlet and outlet pressure ratios of centrifugal impeller and radial turbine measured by experiment are equal to those predicted by BP neural network identification model within the allowable error range.Thus,the accuracy of the BP neural network identification model is verified.