Multi-objective sequential model predictive control for high-power railway induction motor application
Bildiriler > Tam Metin Bildiri · 2024
Yazarlar: Baldomero Araya, Marco Rivera, Carlos Restrepo, PATRICK WHEELER, Emrah ZERDALİ
IET Conference Proceedings, Birleşik Krallık, Nottingham, Uluslararası, Sözlü Sunum, Tam Metin Bildiri, 2024-2024, Scopus, cilt 2024, ss 613-619
Özet
Model Predictive Control (MPC) has demonstrated its effectiveness in several industrial applications, but it grapples with weighting factor (WF) tuning challenges. This paper presents a comparative analysis between conventional MPC and Sequential MPC (SMPC) for an induction motor driven by a three-level neutral point clamped (3L-NPC) inverter. Unlike classical MPC, the SMPC technique does not incorporate a global cost function. Instead, it employs a cascade of simple cost functions, thereby eliminating WFs. Simulation studies were conducted using parameters of a high-power motor for railway applications, incorporating up to four control objectives. The comparison metric employed is the steady-state total harmonic distortion index, with SMPC prevailing in the majority of scenarios. This outcome signifies the potential of achieving commendable motor performance while circumventing the arduous WF tuning process. Additionally, a parameter variation test was conducted, demonstrating the control’s robustness against such testing conditions.