Generalized Dynamic Predictive Control for Nonlinear Systems Subject to Mismatched Disturbances With Application to PMSM Drives

• Published in: IEEE transactions on industrial electronics (1982) Vol. 71; no. 1; pp. 954 - 964
• Main Authors: Dong, Xin, Mao, Jianliang, Yan, Yunda, Zhang, Chuanlin, Yang, Jun
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.01.2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Continuous-time predictive control; Control methods; Disturbances; Drives; Employee welfare; mismatched disturbance; Nonlinear control; Nonlinear systems; Optimization; permanent magnet synchronous motor (PMSM); Permanent magnets; Perturbation; Predictive control; Regulation; robustness and adaptiveness balance; self-tuning receding horizon; Synchronous motors; Trajectory; Vehicle dynamics
• ISSN: 0278-0046; 1557-9948; 1557-9948
• DOI: 10.1109/TIE.2023.3245213

This article investigates a generalized dynamic predictive control (GDPC) strategy with a novel autonomous tuning mechanism of the horizon for a class of nonlinear systems subject to mismatched disturbances. As a new incremental function for the predictive control method, the horizon can be determined autonomously with respect to the system working conditions, instead of selecting a fixed value via experience before, which is able to effectively improve the control performance optimization ability to a certain extent considering different system perturbation levels. To this aim, firstly, a nonrecursive composite control framework is constructed based on a series of disturbance observations via higher-order sliding modes. Secondly, by designing a simple one-step scaling gain update mechanism into the receding horizon optimization, the horizon can be therefore adaptively tuned according to its real-time practical operating conditions. A three-order numerical simulation and a typical engineering application of permanent magnet synchronous motor drive system are carried out to demonstrate the effectiveness and conciseness of the proposed GDPC method.