Abstract:
The demand for electricity has been increased tremendously in the last decade due to the
rapid increase in population and exponential growth. The main source of electricity is the
conventional energy sources i.e., coal, petroleum and natural gas, etc. However, these non renewable energy sources are getting depleted quickly and putting adverse effects on the
environment by the emission of greenhouse gases. Moreover, the current conventional
energy system is also facing difficulties in fulfilling the electricity demands, due to
increasing power outages, coal prices, and the amount of electricity wastage during
transmission through the grids. Therefore, renewable energy sources i.e., solar energy,
wind power, and hydroelectricity have been getting immense popularity in the world and
they have very little effect on the environment. However, solar energy has numerous
advantages over others. Therefore, researchers have been investigating various techniques
to convert solar energy to electricity with maximum efficiency. One of the critical problems
is the complex partial shading, this occurs when the clouds or any other object stops the
light from hitting the panels and this results in two or more peaks that globalize the MPP
and it becomes a nonlinear problem. To solve this problem, in this research work a hybrid
technique based on Perturb and Observe (P&O) and Dragonfly Algorithm (DA) has been
implemented to track maximum power point under both partial shading and complex
partial shading scenarios. The simulations are done on Simulink MATLAB and different
techniques i.e., P&O, DA, Particle Swarm Optimization (PSO), and Cuckoo Search
Algorithm (CSA) are compared with the purposed hybrid technique under four different
irradiances i.e., uniform irradiance, partial shading 1, partial shading 2 and complex partial
shading. In all cases, the results have been compared and it shows that the developed
technique is superior to the compared techniques in terms of transient and power loss.
Similarly, GMPP is tracked faster than any method compared with, because the P&O which
works before DA reduces the search space and time to achieve GMPP as well
