基于伪谱法的滑翔轨迹多级迭代优化策略

Multistage iterative optimization strategy for gliding trajectory based on pseudospectral method

摘要:

伪谱法可实时求解具有高度非线性动态特性的飞行器最优轨迹；以X-51A相似飞行器模型为研究对象,采用增量法与查表插值建立纵向气动力模型,伪谱法与序列二次规划算法求解滑翔轨迹最优控制问题；提出使用多级迭代优化策略,为序列二次规划算法求解伪谱法参数化得到的大规模非线性规划问题提供初值,弥补序列二次规划算法在求解大规模非线性规划问题过程中,出现的初值敏感、收敛速度减慢等问题。通过与传统方法求解出的状态量与控制量仿真飞行状态进行对比,证明了多级迭代优化策略的有效性和高效性,该策略在实际工程应用中取得了良好效果。

Abstract:

The pseudo-spectral method can solve the optimal trajectory of the aircraft with high nonlinear dynamic characteristics in real time. Taking the X-51A similar aircraft model as the research object, the longitudinal aerodynamic model is established by the incremental method and table look-up interpolation. The pseudo-spectral method and the sequential quadratic programming algorithm are used to solve the optimal control problem of gliding trajectory. A multistage iterative optimization strategy is proposed to provide initial values for solving large-scale non-linear programming problems parameterized by pseudo-spectral method, and to compensate for the sensitive initial values and slow convergence of the sequential quadratic programming algorithm in solving large-scale non-linear programming problems. By comparing the state variables and the control variables simulated flight state with the traditional method, the effectiveness and efficiency of the multi-level iterative optimization strategy are proved. The strategy has achieved good results in practical engineering applications.