To address train rescheduling and speed management under section disruptions for urban rail transit, this study constructs a mixed integer nonlinear programming (MINLP) model with decision variables including section running time, section acceleration time, additional dwell time, and whether a train is affected by a disruption, aiming to minimize train traction energy consumption and terminal arrival delay times, providing a coordinated multi-objective optimization framework for restoring train operations in emergency scenarios, and thereby enabling rapid optimization of train rescheduling and energy-efficient speed profiles and ensuring the safe and stable operation of urban rail transit systems. Considering the quadratic non-convex constraints in the model, we reformulate the MINLP into a mixed integer quadratic programming (MIQP) model through Taylor approximation method, enabling effective solutions via commercial solvers. Utilizing operational data from Beijing Subway Yizhuang Line, numerical experiments validate the effectiveness of the MIQP in reducing train terminal delay time and traction energy consumption. This research provides a theoretical basis for rail transit operators to optimize train dispatching and operational control during disruptions. Keywords: urban rail transit; train rescheduling; speed management; energy-efficient train operation; taylor approximation