Abstract: A new augmented model reference adaptive compensation control method was developed， in order to improve the control performance degradation caused by the uncertainty of the multivariable system. Firstly， a basic controller was established with the optimal LQR control law to optimize the performance of the system. And then the optimal LQR basic controller was improved under the feedback control structure to improve the dynamic tracking and anti-interference performance of the system. Finally， the simulation of the controller was carried out with a certain type of double rotor turbo-fan engine. The simulation results showed that the LQR basic controller improved by the augmented model reference adaptive compensation method realizes the tracking compensation to the uncentainty， the tracking of the control command and reaches the desired response performance with control error less than 0.25%， over shoot less than 0.5% and response time less than 1.5s. This improved LQR basic controller satisfies the technical requirements of the aero-engine control system. Meanwhile， this new adaptive compensation method improves the performance of the original basic controller with the uncertainty and ensures that the system is uniformly asymptotically stable.

Key words: Aero engine；Command tracking；Uncertainties；Optimal controller；Augmented MARC

摘要： 针对多变量系统存在不确定性导致的控制性能下降问题，提出了一种基于最优控制律的增广模型参考自适应控制器补偿设计方法。通过采用最优LQR控制律实现系统性能的优化设计从而建立基本控制器，以反馈控制结构为框架，对该最优LQR基本控制器实现增广设计，以改善系统的动态跟踪和抗干扰特性，以双转子航空发动机为对象实现控制器的仿真验证。结果表明:通过增广自适应控制实现了原LQR基本控制器对系统不确定性的跟踪补偿，有效地实现了控制指令跟踪，达到了期望响应性能，控制误差小于0.25%，超调量小于0.5%，且调节时间小于1.5s，符合发动机控制系统技术要求；同时，改善了原基本控制系统在不确定性时的不稳定控制效果，保证了系统一致渐近稳定。

关键词: 航空发动机；指令跟踪；不确定性；最优控制器；增广MRAC