@Article{cmc.2026.080895,
AUTHOR = {Junyi Wang, Jianghai Geng, Jiaqi Liu, Haibin Zhu},
TITLE = {An Intelligent Thermal Monitoring Platform for Manufacturing Workshop Power Distribution Systems},
JOURNAL = {Computers, Materials \& Continua},
VOLUME = {},
YEAR = {},
NUMBER = {},
PAGES = {{pages}},
URL = {http://www.techscience.com/cmc/online/detail/26764},
ISSN = {1546-2226},
ABSTRACT = {In intelligent manufacturing and remanufacturing systems, the thermal safety of the power distribution infrastructure is crucial for ensuring production continuity, equipment reliability, and operational resilience. Traditional temperature monitoring methods often have problems such as high deployment costs, strong environmental sensitivity, or limited physical interpretability in distributed workshop environments. To address these limitations, this study proposes a physically information-driven intelligent thermal color-changing fault identification framework. Based on thermochromic experiments, irreversible color-changing coatings are selected, which are combined with a visual-based computing pipeline for autonomous overheating detection. The framework proposes a thermal fault temperature identification algorithm based on the self-organizing map (SOM) neural network, which improves the generalization ability through RGB channel normalization; multiple random initialization and optimal screening mechanisms are introduced to explore the global optimal solution in the weight space, addressing the issues of sensitive initial weights and easy falling into local minima; the best matching unit (BMU) identification strategy is used to achieve efficient mapping from the RGB feature space to the physical temperature quantity. This algorithm can effectively convert physical color information into a visual temperature field, providing a new idea for thermal fault monitoring and temperature identification of power distribution lines in manufacturing workshops. To evaluate its engineering applicability, a noise resistance experiment with dual scenarios, dual samples, and five noise levels was designed, quantitatively analyzing the absolute error (AE) and F1 score of the algorithm under different noise levels. Laboratory case studies show that under moderate noise interference (<mml:math id="mml-ieqn-1"><mml:mi>σ</mml:mi></mml:math> &lt; 0.04), the algorithm still maintains a high recognition accuracy, verifying its robustness and practical potential in complex industrial site environments. This study embeds the physically constrained material response mechanism into the intelligent visual framework, establishing a new paradigm for computer vision-based manufacturing systems.},
DOI = {10.32604/cmc.2026.080895}
}