Due to the nonlinearity, coupling, and delay during the copper smelting process blowed by enriched oxygen at bottom, a novel two-layer predictive control scheme is presented. The objective of the scheme is to minimize the cost of the material. The two-layer predictive control method achieves the steady-state optimal setpoints and resolves the multivariable dynamic control problems synchronously. Industrial test results show that the two-layer predictive control strategy leads to a significant improvement of control performance, as compared to manual controller within a multiloop framework.