Research On Design And Optimization Of Control In Aero-Engine Transient State

Aero-engine,known as“pearl on the crown of modern industry”,is an important embodiment of science,technology,industry and national defense strength.In 2016,project of aero-engine and gas turbine engine is classified as the first planning of 100 major projects in“13th Five-Year”.Aero-engine control system is an important component of engine system and greatly effects the safety,reliability and economy efficiency of engine.During the research and development of engine control,transient state control takes up nearly 3/4 of total cycle because it should drive the engine with a good dynamic performance over a large nonlinear range,while protect the engine from exceeding its physical limits.So transient state control directly effects engine's safety and life.In this paper,research on control of aero-engine in transient state is carried out.First,FWorks,an integrated digital simulation platform designed for collaborative research and development on FADEC systems for aero-engine,is presented.The internet-based and an open architecture simulation platform,containing baselined numerical models for engines,FADEC systems and components,provides multiple users with flexibility,concurrently and remotely,to construct new or change existing models and control algorithms.Then,a new method for design of transient state control is proposed,and the controller is designed for JT9D engine model.The JT9D component level model is integrated into FWorks using Model Based Design?MBD?method.The modeling method for JT9D is introduced and the embedded code is generated by using“Code Generation”tool in Simulink.Then the C code model is integrated into FWorks to realize real-time simulation.Afterwards,the design method of aero-engine controller is described in detail.First,structure of the controller is presented.Then the“Aero-engine Set-point Controller Design Tool?AESCD?V1.0”is developed for engine model linearization and set-point controller design.Experiment results show that the calculating period of AESCD is less than 1 second,while relative errors between linear model and component level model less than 10^-6.Then the Variable Replacement Method?VRM?is introduced,and the transient state controller is designed based on VRM and PSO algorithm.Simulation results show that compared to the controller in TTECTrA,the accelerating time on ground from idling to take-off decreases by 22.8%.Finally,a rapid prototype based on National Instruments CompactRIO and LabVIEW is built.Simulation results show that the aero-engine operates within its physical limits during the whole accelerating or decelerating,which verifies the effectiveness of controller designed in this paper.