Detailed quantitative comparison of half-bridge modular multilevel converter modelling methods

Guo, Deyang and Rahman, M.H. and Adam, G.P. and Xu, Lie and Emhemed, Abdullah and Burt, Graeme and Audichya, Yash; (2018) Detailed quantitative comparison of half-bridge modular multilevel converter modelling methods. In: The 14th IET International Conference on AC and DC Power Transmission (ACDC 2018). IET, CHN, pp. 1-8.

[thumbnail of Guo-etal-ACDC2018-Detailed-quantitative-comparison-of-half-bridge-modular]
Preview
Text. Filename: Guo_etal_ACDC2018_Detailed_quantitative_comparison_of_half_bridge_modular.pdf
Accepted Author Manuscript

Download (893kB)| Preview

Abstract

This paper presents a detailed comparison of different modelling methods of the half-bridge modular multilevel converter (HB-MMC), namely, switching function, Thevenin equivalent and averaged, considering both MMC implementations (large and reduced number of cells). The theoretical basis that underpins each modelling method are discussed. Offline PSCAD simulations are used to validate user-defined switching function and averaged MMC models against the Thevenin equivalent model provided in PSCAD library for accuracy, considering steady-state and dc fault conditions. Furthermore, the RTDS based real-time simulation results of the user-defined HB-MMC switching function model are validated against the above mentioned offline models, considering steady-state and dc short circuit fault operations. Simulation speed and efficiency of different offline HB-MMC models being studied in this paper are compared. From comprehensive corroboration of different HB-MMC models presented in this paper, it has been found that the averaged, switching function and Thevenin equivalent models produce practically identical results during steady-state and dc faults. In detailed offline and real-time simulation studies where fundamental and harmonic dynamics are of interest, switching function model is found to be faster and computational efficient compared to the Thevenin equivalent model.