Modeling of High-Efficiency Hybrid PV-TEG System with Intelligent Control to Harvest Maximum Energy under Various Non-Static Operating Conditions

Abstract

As the world is moving towards cleaner energy to cater the effects of global warming, the existing renewable energy resources needs to be hybridized with other resources for better output using the same input as before. Photovoltaics-Thermoelectric Generator (PV-TEG) energy system is one such example. The energy from the sun primarily contains two components which carry energy i.e., visible and thermal spectrum, in normal operation of PV energy system, only the visible spectrum is utilized and a considerable amount of energy from the sun is wasted from the thermal spectrum in form of heat. High cell temperature and dynamic temperature spread (DTS) causes current mismatching problem and causes hot spots, resulting in either reduction of efficiency or permanent structural damage due to thermal stresses. The heat which is concentrated at the back of PV panels can be converted into useful energy using series parallel connection of TEG modules resulting in PV panels cooling as well as added energy at the output. For the PV-TEG energy system the controllability aspect is crucial as the main problem lies in the optimization and harvesting of energy from these two sources, the non-linear energy generation nature of the PV and TEG energy systems due to changing conditions i.e., partial shading (PS) and dynamic temperature spread (DTS), makes it hard to attain the full potential of PV and TEG systems using classical/analogue techniques of maximum power point tracking (MPPT). This research work employed the novel implementation Flying Squirrel Search Optimization (FSSO) to find the maximum power point (MPP) for PV-TEG energy system. Compared to already implemented algorithms, FSSO owns the distinctive superiority of simple implementation structure. Meanwhile, its own random and adaptive parameters selection principle greatly boosts the convergence performance which is lacking in all the existing algorithms. To validate the superiority of the results from FSSO, multiple case studies are made to compare the results with the existing promising algorithms such as Particle swarm optimization (PSO), Fruit-Fly optimization (FFO), Perturb and observe (P&O) and Incremental conductance (INC). The simulations confirmed the robustness of FSSO, as it achieves more power and improved tracking time compared to other