Machine Learning-Based Power Quality Improvement of Wind-PV System Using ANFIS

Abstract

The drastic improvements in Energy utilization are rooted in the search for alternative sources for generating Electricity. Due to their advantages for the environment, the integrations of RES into the powers grids, like wind and photovoltaic (PV) systems, has received a lot of attention. The grid’s power quality is challenged by the intermittent and erratic nature of various energy sources. In this research, enhanced Random Forest and ANFIS-based intelligent control algorithms are proposed to reduce the power quality issues in the hybrid PV-integrated Directly Driven Synchronous Generator efficiently. Here, the supervised RF-ANFIS algorithms are acting as a hybrid approach of machine learning techniques that will dynamically allocate the switching sequence for the power converters connected to the grid. Finally, the THD is reduced to less than 2.25% with a time period of 3 ms and upholds voltage stability within allowable bounds. Due to this, the system successfully raises the power factor, creating a steady and dependable power supply. The advancement of more effective and dependable renewable energies integration into the powers grids will ease the transition to a future with resilient and sustainable energy sources.