Aero-engine Control And Verification Based On Model Reference Adaptive Controller

In recent years,with the rapid development of modern industrial technology,aero-engines as the most technologically advanced products in the equipment manufacturing industry have been received extensive attention from industry and academia.Due to the inherent strong non-linearity,uncertainty,time variability,and other characteristics of aero-engines,strict require-ments are put forward for its controller.As a result,the existing products adopt a conservative gain scheduling PID controller,which sacrifices part of the control performance and reduces the economic benefits.Therefore,to improve the control performance of the aero-engine,the research of the multivariable robust adaptive controller is carried out in this paper,combined with hardware-in-the-loop simulation tests to provide feasibility and real-time verification for practical engineering applications.The main work of this paper is as follows:(1)The modeling techniques of the nonlinear mathematical model,the actuator model,and the reference model in the aero-engine control systems are studied.This paper takes the dual-rotor turbofan engines as the controlled object and is modeled by a combined equation group based on volume dynamics,gas dynamics,thermodynamics,and rotor dynamics.At the same time,this paper analyzes the causes of the NaN errors in the model calculation process and gives the solutions.Besides,a simplified model of the actuator is designed.The high-pressure rotor speed and the turbine pressure drop ratio,which are most relevant to the thrust and have the smallest coupling,are selected as the controlled variables to establish an incremental linearization reference model for controller design.(2)A set of universal multivariable controller scheme is designed,including a starting open-loop controller,a bumpless switching controller,an actuator controller,and a main controller.The bumpless switching controller skillfully converts the main controller into an incremental controller through a common integrator,which avoids sudden changes in the system at the con-troller switching time.The actuator controller adds the anti-windup strategy to effectively pre-vent the system output from falling into the saturation region.The main controller adopts the multivariable robust model reference adaptive controller based on an augmented Linear Quadrat-ic Gaussian controller with dead zone and projective operator correction,which effectively im-proves the robustness and adaptability.Furthermore,an anti-windup controller for open-loop stable systems is proposed to avoid undesirable phenomena such as prolonged lag time and in-creased overshoot due to limited input of the model.(3)The numerical simulation test is carried out.Firstly,the parameters of the reference model are obtained by the system identification method.Then,the specific coefficients of the controller are tuned in turn according to the parameter adjusting sequence.Finally,four numer-ical simulation tests,including nominal state,control gain bias,parameter uncertainty,and full working envelopes test,are carried out.The simulation results show that the multivariable con-troller in this paper has excellent robustness and adaptability,which can effectively improve the control performance of the engine.(4)The hardware-in-the-loop test is carried out.Based on the requirement analysis of the hardware-in-the-loop system,the software and hardware design schemes of the system are intro-duced in detail.The hardware part provides both a real-time environment for the model and input and output interface signals for the controller.The software part is a model-based development method and is developed on the basis of the original controller engineering.Finally,through the analysis of the hardware-in-the-loop test results,the effectiveness of the hardware-in-the-loop system,the real-time performance of the controller,and the feasibility of its application in engineering are fully verified.