Advanced Control Strategies for Photovoltaic Power Quality and Maximum Power Point Tracking Optimization

Abstract

This paper introduces an innovative Expectation-Conjugate-Gradient (ECG) approach designed for the management of the interfacing inverter in a grid-connected fixed tilt bifacial Photovoltaic (PV) system. The focus of this algorithm is to address current harmonics issues and elevate power quality. Additionally, the proposed control strategy integrates with an Enhanced Dual second-order generalized integrator phase-locked loop (EDSOGI-PLL), enhanced with a Savitzky-Golay Filter (SGF) to achieve grid synchronization, mitigate voltage harmonics, and estimate symmetrical components during unbalanced grid conditions. The state-of-art Double deep Q-network (DDQN) maximum power point tracking algorithm is introduced, integrating centralized start-up condition computation to stabilize voltage levels at the shared DC-link bus terminal. Key benefits include reduced harmonics, enhanced stability, adaptive control, and minimal computational load. Simulation confirms compliance with IEEE 519 Standards, THD <3%, PF $\approx ~1$ , and reveal a significant reduction in CO2 emissions.

Keywords

• Expectation-conjugate-gradient
• double deep q-network
• bifacial photovoltaic system
• power quality
• maximum power point tracking
• total harmonic distortion