Long Short Term Memory Tracker‐Based Modified DC‐DC Converter for Power Quality Improvement in Grid‐PV Systems under Uniform and Partial Shading Environments

Mohan Kuppusamy,Sujatha Balaraman

DOI: https://doi.org/10.1002/ente.202300648

IF: 4.149

2024-01-15

Energy Technology

Abstract:Proposed model contain three stages: Initially, a dataset is created based on photovoltaic array which contain temperature and irradiance. Then a deeplearning model is designed. Finally an optimization based controller is used to generates a pulse signal for converter. In previous era, coal and oil like non‐renewable resources were mostly used for power generation, thus causing a severe environmental problem as well as high economic. An intelligent maximum power point tracking (MPPT) system was proposed to overwhelm the above‐mentioned impacts, which tracks the maximum generation of photovoltaic (PV) under various atmospheric conditions. Moreover, a modified DC‐DC converter was introduced to low inrush current and switching oscillation. The proposed model working process was split into three steps: first, design the normal PV with a grid‐integrated power system. Second, create a dataset with the aid of a designed PV model, and finally, design an intelligent MPPT controller based on that dataset. The difference among the MPPT and PV voltage was fed to an optimal controller that generates an appropriate pulse signal of the modified DC‐DC converter. The proposed intelligent tracking system was analyzed under various circumstances like static, dynamic and combined shading conditions. The proposed model offers 99.5% efficiency and very low harmonics values like 0.15% current total harmonic distortion (THD) and 0.46% voltage THD. The outcome verifies that the intelligent MPPT approach performs well than that of the currently used controllers in solar power generation and conditioning systems.

energy & fuels