Research On Identification Method Of Aeroengine Closed-loop Control System

Aero-engine is the heart of aircraft.In order to monitor the safe state of aero-engine and ensure the good performance and safe operation of aero-engine,the mathematic model of aero-engine control system should be established.Because of the complex structure of aero-engine,it is very difficult to model,control and diagnose aero-engine with traditional mechanism modeling method.In this paper,recursive least squares identification with forgetting factor and(0.4,0.7)stochastic inertia weight particle swarm optimization identification methods are proposed to establish the mathematical models of aeroengine under three typical operating conditions.The establishment of aero-engine mathematical model by system identification is simple and practical,so it is of great significance in the research of aero-engine closed-loop control system.The main contents of this paper are divided into three parts:(1)based on the DGEN380 engine virtual experiment platform,the excitation signal is designed by estimating the transition time and the maximum operating frequency of the system.The input and output sequence of aeroengine closed-loop control system is collected,and the order of the model is determined by AIC rule.The least square identification method and(0.4,0.7)uniform distribution random inertia weight particle swarm optimization method are used to identify the aeroengine system.Finally,the relative maximum error percentage and relative mean square error percentage are used to verify the accuracy of the model.(2)for the Least Square Identification Algorithm,the recursive least square algorithm with forgetting factor is introduced to estimate the parameters of the slow-moving phase,the cruise phase and the climb phase of the aeroengine closed-loop control system respectively.The simulation results show that the relative maximum error percentage is less than 8% and the relative mean square error percentage is less than 6%.(3)for the particle swarm optimization(PSO)Algorithm,the higher the Inertia Weight Coefficient will be beneficial to the search of global extremum,and the lowerthe inertia weight coefficient can improve the accuracy of the algorithm,a particle swarm optimization(PSO)identification algorithm with(0.4,0.7)uniformly distributed random inertia weight coefficients is proposed to estimate the parameters of aeroengine closed-loop control system in slow-moving phase,cruise phase and climb phase.The simulation results show that the relative maximum error percentage is less than 5.5% and the relative mean square error percentage is less than 3%,compared with the recursive least square algorithm with forgetting factor,the particle swarm optimization Algorithm(0.4,0.7)with uniform distribution and random inertia weight coefficient has higher accuracy and convergence.