Abstract: For the problem that moisture in pneumatic conveying drying is difficult to be stabilized,a dual-feedback control system composed of a soft-sensing model for moisture and an expert controller for nitrogen gas is put forward.The output of the soft-sensing model is obtained by an intelligent coordinator.The optimal value of fuel is searched using genetic algorithm based on the constraints,and the oxygen content in front of rotary-kiln is adjusted with the expert controller for nitrogen gas in order to ensure the ore concentrate not to combust.The least value of fuel and nitrogen gas is obtained simultaneously by the dual-feedback control.The scheme is tested by industrial data,and its average value of the fuel and nitrogen gas is decreased by 1.4% and 0.3% respectively.