Robust-to-Early Termination Model Predictive Control

• 成果类型: Article
• 署名作者: Hosseinzadeh, Mehdi; Sinopoli, Bruno; Kolmanovsky, Ilya; Baruah, Sanjoy
• 署名单位: Washington State University; Washington University (WUSTL); University of Michigan System; University of Michigan; Washington University (WUSTL)
• 刊物名称: IEEE TRANSACTIONS ON AUTOMATIC CONTROL
• ISSN/ISSBN: 0018-9286
• DOI: 10.1109/TAC.2023.3308817
• 发表日期: 2024
• 页码: 2507-2513
• 关键词: Optimization; Predictive control; trajectory; steady-state; Lyapunov methods; Level set; dynamical systems; Barrier function; early termination; limited computing capacity; model predictive control (MPC); primal-dual flow

摘要

Model predictive control (MPC) is a popular control approach to ensure constraint satisfaction, while minimizing a cost function. Although MPC usually leads to very good results in terms of performance, its computational overhead is typically nonnegligible, and its implementation for systems where the computing capacity is limited may be impossible. To address this issue, this technical note proposes a robust-to-early termination MPC. That is, the proposed scheme runs until available time for execution runs out, and the solution, while suboptimal, is guaranteed to enforce the constraints and ensure recursive feasibility despite arbitrary early termination. Also, the closed-loop stability is maintained. Simulations are carried out on an F-16 aircraft to assess the effectiveness of the proposed scheme.