Abstract: The dynamic optimization control strategy for reheating furnaces based on thediscrete-time state space model is investigated in this paper.Due to the complicity of the process dynamics,the control strategy is divided into two parts:steady state zonetemperature set point optimization and its dynamic compensation.Heuristic search technique is adopted in solving the optimization problem while feedback-feedforward principle is applied to dynamic compensation procedure. The proposed algorithm has the advantages of less computation and fast convergence speed.The dynamic compensation is capable of overcoming the various disturbances from the process.The implementation of the proposed strategy shows a great su-periority to the original control scheme,included energy saving (9%),steel loss decreasing (0.18%) and better product quality,thus producing over 1 million Yan/yeareconomic benifit.