A computational study published in Results in Engineering demonstrates that reconfiguring solar photovoltaic arrays using an algorithmic torus puzzle strategy significantly enhances power extraction and suppresses mismatch losses under non-uniform illumination. Conducted by an engineering research team, the investigation evaluated nine-by-nine photovoltaic arrays wired in a Total-Cross-Tied topology under diverse partial shading conditions. By treating module placement as a cyclical sliding puzzle, the computational approach redistributes shaded modules across array rows to balance current generation. In simulations across three distinct partial shading scenarios, the proposed torus puzzle framework achieved global maximum power points reaching up to 77.4 ImVm, outperforming traditional Total-Cross-Tied layouts which reached only 65.7 ImVm under identical irradiance distributions. The numerical findings confirm that systematically rearranging modules using predefined geometric translation rules enables solar installations to recover significant energy that would otherwise be dissipated as unusable heat across shaded solar cells.
The fundamental motivation behind developing array reconfiguration models stems from the severe operational inefficiencies induced when photovoltaic arrays experience non-uniform solar radiation. In commercial and residential installations, obstructions such as adjacent buildings, overhanging tree canopies, and drifting clouds frequently cast irregular shadows over portions of solar arrays. When sunlight distribution across an array becomes uneven, individual solar cells operate at divergent irradiance levels, generating unequal current outputs. This electrical discrepancy disrupts uniform current flow, giving rise to multiple local peaks along the electrical power-voltage characteristic curve and creating a complex global maximum power point tracking challenge. Furthermore, localized current mismatches force shaded cells into reverse-bias operation, triggering localized thermal stress commonly designated as hot spots. Over extended operational periods, these persistent hot spots degrade overall power conversion efficiency, accelerate module aging, and can inflict permanent physical damage upon shaded semiconductor elements, highlighting the technical necessity for effective mitigation mechanisms.
Engineers classify mitigation strategies for partially shaded solar arrays into passive and active technical solutions, each presenting distinct economic and operational trade-offs. Passive techniques rely predominantly on bypass diodes integrated across module groups and hardwired interconnection topologies engineered to diminish mismatch losses. Standard topological layouts include Series-Parallel, Bridge-Link, Honey Comb, and Total-Cross-Tied configurations, with extensive prior literature demonstrating that Total-Cross-Tied networks deliver superior maximum power point tracking performance relative to alternative fixed connections. Conversely, active mitigation methodologies focus on dynamic power electronics or structural reconfiguration, commonly categorized into multi-tracker converter systems, dedicated micro-converter installations, and switch-based array reconfiguration. While multi-tracker converters and micro-converters enhance operational energy capture by managing power at decentralized granular levels, deploying dozens of individual converters dramatically escalates capital expenditures and maintenance overhead. Array reconfiguration instead modifies module interconnection schemes, offering a substantially more cost-effective pathway to mitigate mismatch losses and optimize harvestable energy within Total-Cross-Tied arrays.
Within photovoltaic array reconfiguration, methodologies diverge into dynamic switching architectures and static predetermined arrangements. Dynamic reconfiguration systems perform real-time adjustments of connections between solar modules using extensive sensor networks, microcontroller units, and elaborate switching matrices to continuously adapt to wandering shadows. Although dynamic systems successfully adapt to shifting irradiance, the prerequisite electrical hardware introduces substantial costs, parasitic switching losses, and control vulnerabilities. In response, static reconfiguration methods have emerged as an attractive, highly reliable alternative characterized by rapid response speed, operational simplicity, and the complete elimination of complex switching matrices and real-time irradiance sensors. Over recent years, researchers have formulated numerous static geometric models, including the 8-Queen’s arrangement, Magic Square puzzles, SuDoKu schemes, Triple X puzzles, Skyscraper puzzles, Odd-Even configurations, and Knight’s tour algorithms. However, established static models exhibit notable engineering constraints, frequently struggling with scalability beyond small nine-by-nine layouts, incurring excessive wiring complexities on irregular boundaries, or demanding intensive mathematical computations that complicate practical field application.
To overcome these geometric and computational bottlenecks, the authors developed a novel static reconfiguration method modeled after the mechanics of a two-dimensional torus sliding puzzle. In this mathematical formulation, an array comprising m rows and n columns is conceptualized as a toroidal grid containing m multiplied by n distinct module pieces, numbered sequentially from one to mn. The initial configuration positions each module according to integer row and column coordinates derived from arithmetic division and modulus functions. The algorithm alters module positions using four fundamental cardinal operations designated as North, South, East, and West. The vertical North and South operations cyclically shift elements within a designated column upward or downward, while the horizontal East and West operations cyclically rotate elements along a specified row. Because boundary conditions wrap continuously like a torus, elements exiting one edge smoothly reenter from the opposite side. To optimize computational tractability, consecutive identical operations are treated as single composite steps, enabling efficient algorithmic navigation through candidate spatial configurations.
The systematic implementation of the torus puzzle strategy operates through an established five-stage computational workflow designed to identify the optimal module arrangement for any given shading scenario. In the initial stage, researchers define baseline matrix dimensions representing physical module coordinates alongside the non-uniform irradiance distribution across the array surface. During the second stage, the algorithm generates candidate configurations by executing permitted cyclic transformations along rows and columns, systematically shifting module assignments without altering the total module count. In the third stage, the electrical response of each candidate configuration is computed using standard photovoltaic diode equations, generating complete current-voltage and power-voltage curves to locate the resulting global maximum power point. The fourth stage subjects candidate configurations to defined performance selection criteria, iteratively refining spatial placements until predefined convergence conditions are satisfied. Finally, the fifth stage executes comparative benchmarking, directly contrasting the selected torus configuration against standard Total-Cross-Tied baselines and established puzzle-based methodologies under identical environmental parameters to verify output improvements.
To validate the effectiveness of the torus framework, the researchers established a simulation testbed using MATLAB version 2018 running on a computer equipped with an Intel Core i7 processor and 16 gigabytes of random-access memory. The research team benchmarked the proposed method against standard Total-Cross-Tied connections, Magic Square arrangements, Competence Square puzzles, Skyscraper puzzles, and Improved SuDoKu schemes. In the first simulated case study, an interconnected nine-by-nine Total-Cross-Tied array was subjected to a localized three-by-three shading pattern where irradiance levels dropped sharply to 500 Watts per square meter and 300 Watts per square meter across specific modules. Under these non-uniform conditions, standard Total-Cross-Tied wiring produced a global maximum power point of only 61.2 ImVm, as multiple shaded rows suffered severe current suppression. Competing puzzle arrangements, including Magic Square, Competence Square, Skyscraper, and Improved SuDoKu models, all plateaued at 69.3 ImVm. In contrast, the torus puzzle approach successfully dispersed shading across different electrical rows, achieving a superior global maximum power point of 72.9 ImVm.
The research team further examined system performance across two additional shading distributions to test the robustness of the torus puzzle model under divergent irradiance geometries. Case two evaluated a nine-by-nine array exposed to an asymmetric five-by-two shading footprint, with solar irradiance reduced to 300 Watts per square meter across columns one through three of rows six and seven, and 400 Watts per square meter across columns four and five. The conventional Total-Cross-Tied layout generated 63 ImVm, while Improved SuDoKu yielded 65.7 ImVm and both Magic Square and Skyscraper methods attained 70.2 ImVm. The torus puzzle configuration outperformed all alternatives, attaining 72.9 ImVm by equalizing current contributions across rows. In the third case study, a severe three-by-three shading pattern was positioned across rows seven through nine, with irradiances diminished to 500 Watts per square meter in columns two and three, and 400 Watts per square meter in column four. Total-Cross-Tied wiring yielded 65.7 ImVm, Magic Square reached 68.8 ImVm, and Skyscraper achieved 74.7 ImVm. Demonstrating superior optimization, the torus method achieved 77.4 ImVm.
Beyond evaluating peak power extraction, the research study conducted comprehensive quantitative assessments across several standardized electrical performance metrics, specifically analyzing the Fill Factor, Power Efficiency, and Mismatch Losses. The Fill Factor measures the quality of solar array electrical output by comparing maximum extractable power against the product of open-circuit voltage and short-circuit current. Power efficiency quantifies the percentage of total incident solar energy successfully transformed into usable electrical wattage. Mismatch loss mathematically captures the percentage disparity between theoretical maximum power reachable under uniform illumination and the global maximum power point realized under partial shading conditions. Across all three investigated shading cases, the torus puzzle strategy consistently established superior Fill Factor values and achieved the lowest mismatch losses recorded in the investigation. Specifically, the torus methodology restricted mismatch losses to 3.45 percent in case one, 3.65 percent in case two, and 3.23 percent in case three. These low loss percentages confirmed the model’s distinct ability to balance current outputs across parallel-connected circuit branches.
Despite the compelling performance advantages demonstrated across all three simulated scenarios, the authors explicitly contextualized their findings within the defined boundaries of computer modeling, noting that physical experimental testing remains necessary. The entire investigation was executed within a software simulation environment, meaning physical phenomena such as parasitic cable resistance, connector degradation, and dynamic thermal gradients were not experimentally measured. In physical field installations, additional wiring lengths and interconnecting switches inherently introduce resistive power dissipation governed by the formula Ploss equals I squared R, which could partially erode net electrical power gains achieved through module rearrangement. Furthermore, hardware acquisition costs, switching losses, and control complexity require thorough empirical validation under real outdoor operating conditions. To advance practical deployment, the authors developed microcontroller code capable of orchestrating array reconfiguration based on the torus algorithm. Future research efforts will focus on constructing physical hardware prototypes under controlled shading environments, directly quantifying switching losses, installation costs, and long-term operational reliability in field-scale solar installations.
Subject of Research: Photovoltaic array reconfiguration under partial shading conditions using a torus puzzle model
Article Title: Photovoltaic arrays reconfiguration for maximum energy generation under partial shade conditions using a novel torus puzzle approach
Article References: Moradzadeh, A., Rezazadeh, S., Zare, K., Abapour, M., Lin, S., & Muyeen, S. M. (2026). Photovoltaic arrays reconfiguration for maximum energy generation under partial shade conditions using a novel torus puzzle approach. Results in Engineering, 32, Article 114199. https://doi.org/10.1016/j.rineng.2026.114199
Image Credits: AI Generated
DOI: 10.1016/j.rineng.2026.114199
Keywords: photovoltaic arrays, partial shading conditions, torus puzzle approach, total-cross-tied configuration, array reconfiguration, maximum power point tracking, global maximum power point, mismatch losses, fill factor, power efficiency, static reconfiguration, solar module rearrangement, combinatorial optimization, irradiance distribution, bypass diodes, hot spot mitigation, matlab simulation
Cite Scienmag News
Denise Maddox. (November 18, 2026). Torus Puzzle Method Improves Solar Array Power Output Under Partial Shading. Scienmag. https://scienmag.com/torus-puzzle-method-improves-solar-array-power-output-under-partial-shading/
Denise Maddox. "Torus Puzzle Method Improves Solar Array Power Output Under Partial Shading." Scienmag, 18 November 2026, https://scienmag.com/torus-puzzle-method-improves-solar-array-power-output-under-partial-shading/. Accessed 18 November 2026.
Denise Maddox. "Torus Puzzle Method Improves Solar Array Power Output Under Partial Shading." Scienmag. November 18, 2026. https://scienmag.com/torus-puzzle-method-improves-solar-array-power-output-under-partial-shading/








