Design of a hybrid maximum power point tracker for partial shading conditions in photovoltaic systems

As the demand for energy continues to increase globally, the reserves of fossil-based energy sources are rapidly depleting. This situation has led us to utilize renewable energy sources. Solar energy is one of the primary renewable energy sources and is converted into electrical energy through photovoltaic (PV) panels. PV panels are technologies primarily made of semiconductor materials that directly convert sunlight into electrical energy. With their long service life and ease of installation, PV systems offer a reliable solution to meet energy demands. The performance of PV systems is influenced by several environmental and physical factors, including the amount of irradiance, the angle at which sunlight strikes the panels, ambient temperature, shading, and panel surface cleanliness (e.g., dirt, dust). The characteristic curve of PV panels is nonlinear, and the system can transfer maximum power to the load at only one point, which varies depending on environmental conditions. To extract maximum power from the system, this dynamic point must be continuously tracked, and the system must be operated at this point. To achieve this, Maximum Power Point Tracking (MPPT) systems have been developed. Shading is one of the primary factors negatively affecting the performance of PV panels and is typically caused by clouds, trees, utility poles, or other environmental obstructions. Depending on the type and distribution of shading on the panel, power losses may occur. Under partial shading conditions, shaded areas can absorb the power generated by other sections, potentially leading to hot spot formation. In such cases, bypass diodes connected in parallel with the panels activate to protect the system. The use of bypass diodes complicates the panel's characteristic curve, resulting in the emergence of multiple peak points. Among these peaks, the one providing the highest power is defined as the Global Maximum Power Point (GMPP), while the others are referred to as Local Maximum Power Points (LMPPs). Conventional MPPT algorithms often fall short in identifying the GMPP under partial shading conditions, getting stuck at local maximum power points. Methods such as Perturb and Observe (P&O) and Incremental Conductance (IC) have advantages like low cost and ease of implementation but fail to perform well under complex shading conditions. Various improvement efforts have been made to address the shortcomings of traditional methods, particularly the issue of multiple peak points under partial shading conditions. Examples include developing dynamic step sizes instead of fixed step sizes for the P&O and IC methods. Skip-based methods specifically designed for partial shading conditions are also among these efforts. These methods are combined with traditional approaches to create hybrid structures that address the deficiencies of traditional methods under partial shading. Examples of such methods include the 0.8Voc and SSJ techniques. In this study, some of the traditional and skip-based methods found in the literature are explained, and a new hybrid MPPT algorithm has been developed to enhance the performance of PV systems under partial shading conditions. The proposed hybrid MPPT algorithm consists of three main components. Firstly, the Perturb and Observe (P&O) method ensures the system's operation under regular irradiance conditions. Secondly, the Partial Shading Detection Method identifies whether sudden power changes are caused by partial shading. Lastly, a new Global MPPT algorithm is activated when partial shading is detected. A modified structure has been added to the Partial Shading Detection Method to prevent unnecessary scanning at low voltages, thereby reducing power losses and improving scanning time. Furthermore, the new Global MPPT algorithm has been designed to detect the global maximum power point (GMPP) more quickly and effectively under partial shading conditions, particularly in long PV systems. The new Global MPPT algorithm is fundamentally inspired by the 0.8Voc method. While the skip-based methods studied are effective in PV systems with a small number of series-connected modules, their detection time for the global maximum power point increases significantly as the number of series-connected panels in the system grows. This adversely affects the efficiency of large PV arrays, causing delays in the tracking process. In this context, the designed Global MPPT algorithm aims to eliminate this disadvantage with its structure. The proposed algorithm combines midpoint and sequential scanning methods and selects the most appropriate one to optimize system performance. Thus, the scanning time is reduced, enabling the MPPT system to locate the GMPP more rapidly, even under partial shading conditions, thereby improving efficiency in large PV systems. A PV system designed to implement the proposed hybrid MPPT algorithm was modeled. This system consists of a PV array, a DC-DC buck boost converter, a load, and the MPPT algorithm. Simulation studies were conducted using MATLAB/SIMULINK software. The performance of the proposed hybrid MPPT algorithm was simulated under different atmospheric conditions (e.g., shading, sudden irradiance changes) and various partial shading scenarios. Its tracking capability was analyzed in detail. Simulation results demonstrated that the algorithm accurately tracks the GMPP and enhances system efficiency under partial shading conditions. Particularly, the prevention of unnecessary scans significantly reduced the tracking time. The proposed algorithm was compared with the Incremental Conductance and Skipping Adaptive P&O algorithms, and it was found to outperform them in terms of tracking accuracy, speed, and efficiency. While the Incremental Conductance (IC) method, one of the traditional approaches, failed to track the global maximum power point in many partial shading conditions, the proposed hybrid method and the Skipping Adaptive P&O method successfully tracked the GMPP in all scenarios. Skipping Adaptive P&O method detected the global maximum power point in 10–21 sampling steps (0.25–0.525 seconds) across seven different partial shading scenarios, whereas the proposed hybrid MPPT method achieved this in 3–9 sampling steps (0.075–0.225 seconds). These findings reveal that the proposed algorithm can detect the GMPP faster than the Skipping Adaptive P&O method. In conclusion, the proposed hybrid MPPT algorithm has the potential to enhance system efficiency by rapidly and accurately detecting the global maximum power point, particularly in long PV systems, by reducing scanning time. To further evaluate and improve the algorithm, it can be implemented on real-time hardware platforms in the future, and its practical performance can be assessed.