Finite Control Set Model Predictive Control is an effective technique which attracted attention in the latest years thanks to its fast dynamic response and the fact that it does not require a modulation. However, when applied to multilevel converters, it requires a large amount of calculations that may affect implementability. On the other hand, Neural Networks are well known Machine Learning techniques that can be efficiently implemented in real-time applications thanks to their massive parallelism capability. This work proposes a novel Finite Control Set Model Predictive Control approach based on Neural Networks with reduced computational time complexity for a Cascaded H-Bridge Static Synchronous Compensator. Simulation results for a nine-level system are presented and a simulative comparison between our approach and classical methods for FCS is provided, showing that very similar performance can be achieved with strong reduction of the computational complexity.

A Neural Network Approach for Efficient Finite Control Set MPC of Cascaded H-Bridge STATCOM

Simonetti F.
Methodology
;
D'Innocenzo A.
Supervision
;
Cecati C.
Supervision
2021

Abstract

Finite Control Set Model Predictive Control is an effective technique which attracted attention in the latest years thanks to its fast dynamic response and the fact that it does not require a modulation. However, when applied to multilevel converters, it requires a large amount of calculations that may affect implementability. On the other hand, Neural Networks are well known Machine Learning techniques that can be efficiently implemented in real-time applications thanks to their massive parallelism capability. This work proposes a novel Finite Control Set Model Predictive Control approach based on Neural Networks with reduced computational time complexity for a Cascaded H-Bridge Static Synchronous Compensator. Simulation results for a nine-level system are presented and a simulative comparison between our approach and classical methods for FCS is provided, showing that very similar performance can be achieved with strong reduction of the computational complexity.
978-1-6654-3554-3
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11697/186768
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 0
social impact