{"id":"arxiv/2604.22414","url":"https://arxiv.org/abs/2604.22414","tags":["math.OC","math.NA"],"score":0,"title":"Computational Control of Nonlinear Partial Differential Equations Using Machine Learning","domain":"math.OC","source":"arxiv","authors":["Maximilian Kurbanov","Minh-Nhat Phung","Minh-Binh Tran"],"reviews":0,"subject":"math.OC","summary":"The numerical reconstruction of controls for nonlinear partial differential equations (PDEs) remains a challenging and relatively underdeveloped problem, despite the extensive literature on controllability theory. In this work, we introduce an operator-decomposed physics-informed neural network framework, called WeightedPINN, for approximating controls in nonlinear PDE settings. The method is designed for both internal and bilinear control problems and incorporates the governing equation, boundary and initial conditions, and terminal control constraints directly into the training objective.…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-04-24","reactions":{}}