A Novel Non-Uniform Control Vector Parameterization Approach with Time Grid Refinement for Flight Level Tracking Optimal Control Problems.

Ping Liu,Guodong Li,Xinggao Liu,Long Xiao,Yalin Wang,Chunhua Yang,Weihua Gui

DOI: https://doi.org/10.1016/j.isatra.2017.12.008

IF: 7.3

2017-01-01

ISA Transactions

Abstract:High quality control method is essential for the implementation of aircraft autopilot system. An optimal control problem model considering the safe aerodynamic envelop is therefore established to improve the control quality of aircraft flight level tracking. A novel non-uniform control vector parameterization (CVP) method with time grid refinement is then proposed for solving the optimal control problem. By introducing the Hilbert-Huang transform (HHT) analysis, an efficient time grid refinement approach is presented and an adaptive time grid is automatically obtained. With this refinement, the proposed method needs fewer optimization parameters to achieve better control quality when compared with uniform refinement CVP method, whereas the computational cost is lower. Two well-known flight level altitude tracking problems and one minimum time cost problem are tested as illustrations and the uniform refinement control vector parameterization method is adopted as the comparative base. Numerical results show that the proposed method achieves better performances in terms of optimization accuracy and computation cost; meanwhile, the control quality is efficiently improved.