增广预测模型的航空发动机多变量约束预测控制

针对模型不匹配导致的模型预测控制性能下降的问题，提出了一种基于增广预测模型的航空发动机多变量约束预测控制器设计方法。在现有发动机状态空间模型基础上，将指令跟踪误差与系统状态的变化量增广为状态向量，设计增广预测模型以消除稳态跟踪误差，以控制量所需能量与模型预测输出误差最小为目标，利用带约束的序列二次规划（SQP）算法在线滚动优化控制变量。通过某型涡扇发动机非线性部件级模型的包线内不同状态下仿真分析，结果表明，控制系统无稳态误差，调节时间<2s，有效提高了发动机控制品质，实现了对输出量的限制管理。

Multivariable Constrained Model Predictive Control ofAero-Engine Based on Augmented Predictive Model

To solve the performance degradation of model predictive control caused by model mismatch,a design method of multivariable constrained predictive controller was proposed based on the augmented predictive model for aero-engines. Based on ·the existed state space model, the command tracking errors and variations of system states were augmented as state vector to eliminate the tracking errors. Then, aiming at taking the energy required by control variables and the predictive tracking errors to minimum, the receding horizon optimization of the control variables can be realized online using sequential quadratic programming method. Finally, the controller was applied to the nonlinear component-level-model of a certain turbofan engine where various engine states in full envelope were examined and analyzed. Simulation results show that the settling time of the control system is less than 2 seconds and the steady state error is zero. The quality of engine control is improved effectively and the constraint management of output is achieved.