Abstract: Scheduling steel plate hot rolling is a typical single-machine scheduling problem subject to multiple constraints including precedence constraints and limited waiting time and finite buffer capacity. To describe the steel plate hot rolling scheduling problem with an AON (Activity-on-node) network, an equilibrium principle is formulated for single-machine scheduling problems on the AON networks. Subsequently, a nonlinear constrained optimization model is built for steel plate hot rolling scheduling problems, which maximizes utilization of the rolling mill and is solved by software LINGO. Numerical computations indicate that the hot rolling scheduling method is more effective than the heuristic methods available in reaching optimal production schedules that may significantly improve production stability and efficiency.