A Boundary-Aware PINN Framework for Monkeypox Transmission Dynamics Using Chebyshev-Distributed Collocation
摘要
Monkeypox, a zoonotic infectious disease, has emerged as a major public health concern in recent years, particularly due to its capacity for human-to-human and environmental transmission. This study introduces a novel eight-compartment SVEAIHRC model that incorporates direct, asymptomatic, and environmental transmission pathways along with vaccination and hospitalization effects to better capture the dynamics of mpox spread. The model is analytically investigated to ensure positivity, boundedness, and stability of its solutions, and the basic reproduction number is derived using the next-generation matrix approach. To address the limitations of conventional computational approaches, a modified Physics-Informed Neural Network framework is developed, embedding physical constraints into the training process and employing Chebyshev-based clustering to enhance boundary accuracy. Quantitative validation demonstrates that the proposed method achieves excellent agreement with reference data, with average errors across compartments remaining very low, RMSE