MNFIS and other soft computing based MPPT techniques: A comparative analysis

Abstract

Maximum Power Point Tracking (MPPT) is the process of searching the voltage space for the optimal power generation and tracking the optimum as it changes. This paper presents a performance analysis of soft computing algorithms applied to this endeavor and a deployment recommendation based on performance goals. Specifically, fuzzy logic (FL) and artificial neural networks (ANN) were tested with direct and indirect converter control and compared against multiple metrics for fitness. Along the way a novel algorithm was also developed, deemed the Modified Neuro-Fuzzy Inference System (MNFIS). This algorithm incorporates the strengths of both FL and ANN MPPT while mitigating the weaknesses of either.