부산대학교 도서관

Three-level T-type converters $ {(3L} {T}^{ {2}} {C)}$ with LCL filters are widely used in distributed generation (DG) systems because of their smaller size and better harmonic attenuation. The $ {3L} {T}^{ {2}} {C}$ have many control objectives such as neutral-point (NP) voltage balancing, resonance damping, and grid current tracking. Therefore, model predictive control (MPC) with excellent dynamic characteristics and multiobjective control capability has become the most common control method. However, MPC control relies on accurate mathematical models, and the existing integer order modeling ignores its fractional-order characteristics. The reflected physical characteristics are not sufficient, so an accurate model is needed to improve the control performance. In this article, we propose a fractional-order state-space model based on the Grünwald–Letnikov (GL) definition and investigate a fractional-order sequential model predictive control (FOSMPC). First, the GL definition is used to establish the state-space model of the fractional-order grid-connected converter to realize the discretization processing and numerical calculation of the fractional-order system. Second, the multiobjective lexicographical optimization method is used to realize the SMPC to balance the NP voltage and grid-connected current reference tracking. Finally, the effectiveness of FOSMPC in controlling fractional-order $ {3L} {T}^{ {2}} {C}$ was experimentally demonstrated.