The DC-DC converter is more significant for the integration of photovoltaic (PV) and battery. The design of the power converter is high complexity, because of the nonlinear characteristics, like supply voltage variations, switches, load current variations, and parasitic component fluctuation. This article proposes an efficient approach for increasing the maximum power of PV system and system efficiency. The proposed approach is the consolidation of Random Forest Algorithm (RFA) and Mexican Axolotl Optimization (MAO), hence it is called RFAMAO approach. The proposed approach utilized the single-input and dual-output boost converter to implement the maximum power point tracking of the system that improves the lesser voltage input power source, controllable higher voltage direct current bus and middle voltage output terminals. RFA is used to deal with the uncertainty caused by climate change. MAO is utilized to tune the control parameter of the system. The voltage-mode controlling strategy among the PV, battery, and standalone DC load depending on Digital Signal Processing and Control Engineering (DSPACE) is utilized for uninterruptible power flow management and constant DC load voltage maintenance. The proposed approach is implemented in MATLAB/Simulink, and its efficiency is compared with existing approaches. The oscillation around maximum power point (MPP) for case 1 of RFAMAO, modified shuffled frog leaping algorithm with fuzzy logic (MSFLA-FLC), fuzzy logic, and perturb and observe (P&O) is 2.69, 7.31, 20.54, and 29.12 W. The oscillation around MPP for case 2 of RFAMAO, MSFLA-FLC, fuzzy logic, and P&O is 2.57, 8.34, 18.56, and 27.65 W.
Read full abstract