Adaptive Network-Based Fuzzy Inference System (ANFIS) Applied to Inverters: A Survey

In recent years, adaptive network-based fuzzy inference systems (ANFIS) applied in various engineering fields have demonstrated their ability to combine the learning capability of artificial neural networks and the representation of fuzzy inference systems, achieving promising results, and attracting the research community's interest. This article provides a valuable overview of ANFIS applications. Focusing on the field of electronic engineering, in particular for power electronics applications, the use of ANFIS has proven to be a very effective and reliable tool for applications requiring maximum power point tracking and power quality improvement. For this reason, this article presents an analysis of the review of 59 papers published between 2013 and 2023, of particular interest the applications of ANFIS for power quality improvement and harmonic content decrease in multilevel converters. This article applied a methodology for the systematic literature review that included identification, selection, eligibility, and inclusion criteria that allowed the selection of publications from the various relevant journals and databases consulted. Subsequently, the surveyed publications were analyzed and classified, highlighting the applications, characteristics of the ANFIS employed, their training algorithms, and the platforms or digital cards used. After analyzing the information from the surveyed papers, it can be concluded that using ANFIS has obtained promising results in improving inverter performance and power quality indexes in the face of variations. ANFIS offers flexibility and adaptability by performing adaptive and real-time control by continuously adjusting the inverter operation.