Advanced Mitigation Strategies for Partial Shading in Solar Photovoltaic Systems: A Review of Topologies, Reconfigurations, and Metaheuristic MPPT

The global energy landscape is currently undergoing a transformative shift necessitated by the rapid depletion of non-renewable repositories such as coal, oil, and natural gas. Driven by the escalating demands of industrialisation and exponential population growth, particularly in developing nations, the energy gap between generation and demand continues to widen. Global efforts have subsequently pivoted toward renewable energy sources (RESs) to mitigate the environmental deterioration and carbon dioxide emissions associated with global warming, with many nations pledging to reach Net Zero by 2050. Among these technologies, solar photovoltaic (PV) systems are increasingly prioritised due to their abundance, noise-free operation, minimal maintenance requirements, and their adaptability for installation in confined urban spaces like rooftops. Power generation from RESs not only preserves billions of barrels of crude oil but is essential for establishing sustainable energy security. Despite the proliferation of PV technology, its operational efficiency remains inherently limited by its non-linear output characteristics and a heavy dependence on varying environmental conditions, specifically solar irradiance and cell temperature. While the peak solar density received at the earth's surface is approximately 1000 W/m², the actual power extracted is often much lower due to the PV cell's sensitivity to atmospheric fluctuations. A significant technical hurdle is the occurrence of Partial Shading Conditions (PSCs), which arise from moving clouds, shadows from adjacent buildings or trees, dust accumulation, and even debris such as bird droppings. These factors distract the rate of incident photons, which directly impedes the liberation of electrons within the n-type material and the subsequent flow to the load. Under PSCs, the modules within an array receive non-uniform irradiation levels, causing a mismatch in current generation. Shaded cells produce significantly less photon current than their unshaded counterparts, leading them to become reverse-biased and act as electrical loads rather than generators. This phenomenon results in substantial mismatching power losses (ML) and creates a high risk of hot-spot heating, which can cause irreversible physical damage to the module structure. To protect the array, bypass diodes are typically integrated; however, their activation introduces further complexity by transforming the standard single-peak Power-Voltage (P-V) curve into a non-linear landscape. This multi-modal curve contains several Local Maximum Power Points (LMPPs) and only one absolute Global Maximum Power Point (GMPP). Research indicates that as the number of modules receiving different insolations increases, the number of stairs in the I-V curve and the count of local optima in the P-V curve rise proportionally. The evolution of research into mitigating these losses has transitioned through several distinct technical paradigms. Initially, efforts focused on classical Maximum Power Point Tracking (MPPT) logic, such as Perturb and Observe (P&O) and Incremental Conductance (IC). While these methods are computationally simple, they are inherently flawed under PSCs as they tend to oscillate around the nearest power peak and frequently become trapped at LMPPs, leading to wasted energy. Consequently, the focus shifted toward hardware-level interventions, exploring various interconnection topologies like Series-Parallel (S-P), Total Cross-Tied (TCT), Bridge-Link (BL), and Honeycomb (HC). Comparative analyses have established that the TCT configuration generally provides the most resilient output and highest maximum power extraction across diverse shading patterns. More recently, the research domain has advanced toward computational intelligence and bio-inspired metaheuristics to navigate non-convex search spaces. Algorithms mimicking the natural behaviours of salp swarms (SSA), bats (BA), and fireflies (FA) have demonstrated superior tracking accuracy and faster convergence speeds than conventional methods. For example, the Salp Swarm Algorithm utilizes a direct GBest technique to move particles straight toward potential targets, skipping unnecessary search regions and identifying the GMPP in as little as 0.22 seconds. Similarly, the Bat Algorithm employs echolocation logic to provide reliable tracking without steady-state oscillations. Parallel to these software advancements, puzzle-based reconfigurations using Sudoku, Latin Square, and Futoshiki logic have emerged as a means to physically re-allocate modules for optimal shade dispersion. These methods, such as the Latin Square (LS-TCT) approach, aim to equalise row currents and have been validated to reduce mismatch losses by up to 1787 W in specific scenarios. This review paper provides a comprehensive and systematic synthesis of these contemporary hardware and software strategies. By critically evaluating the effectiveness of array interconnection schemes, dynamic reconfiguration algorithms, and hybrid metaheuristic trackers, this work contributes a clear benchmarking framework for enhancing energy conversion rates. This investigation is vital for selecting optimal PV schemes that can extract maximum power under dynamic operating conditions, ultimately bridging the gap between theoretical generation and practical demand in the pursuit of global energy sustainability.