多变量多目标涡轴发动机最优加速控制

摘要

Based on the dynamic mathematical model of a turboshaft engine with variable inlet guide vane compressor, a Feasible Sequential Quadratic Programming （FSQP） algorithm is applied to the study of multi-variable and multi-object optimal acceleration control through linear weight method. Simulation results show that the proposed optimal control algorithm is feasible and improves dynamic performance of turbo-shaft engine. The results also illustrate that the algorithm shortens acceleration time and makes engine per-formance better with proper weights of multi-object optimization.%基于压气机几何可调的自由涡轮式涡轴发动机部件级动态数学模型，利用可行性序列二次规划（FSQP）算法对多优化目标进行线性加权，提出了多变量多目标的涡轴发动机最优加速控制算法。仿真结果表明，提出的优化算法改善了涡轴发动机加速过程的品质；选择恰当权重的目标函数能大大缩短加速时间，进而改善发动机的动态性能，进一步发挥发动机的性能潜力。