Robust Multiple Model Predictive Control for Ascent Trajectory Tracking of Aerospace Vehicles
Rui Cao, Yanbin Liu, Yuping Lu
Abstract
For aerospace vehicles with uncertainty, strong nonlinearity, and high-performance requirements, the model predictive control (MPC) method has attracted many researchers’ attention with its unique advantages. Although this method has a good effect, it is computationally expensive, especially for the nonlinear MPC. In addition, it cannot guarantee the stability performance requirements and limited flexibility in dealing with large initial errors. To overcome these limitations, this article is implemented to obtain a control scheme that can handle uncertainty, initial deviation, and has less online calculation. To ensure the stability requirements, the guardian maps (GM) theory is combined with multimodel predictive control, and a switching rule based on GM is proposed. Then, fully considering the characteristics of linear matrix inequalities used, the idea of “offline design and online optimization” is adopted to solve the control law, which reduces the burden of online computing. Furthermore, in the case of large initial errors, a transient trajectory design method is established to reconstruct the trajectory online, and to modify the tracking command. Simulation results verify the effectiveness of the designed control scheme.