针对炼钢除尘系统的复杂过程提出了智能控制方法。由基本控制器给定一个基本输出值，通过神经网络预测冶炼不同阶段烟气量变化趋势，并以此来改变系统的工作点，及时调整风机转速。采用Ｆｕｚｚｙ推理的方法补偿布袋过滤器对系统的影响。实际应用表明，该方法可获得很好的除尘效果，降低除尘系统的运行成本。 Aiming at the complex process of steelmaking dust removal system, an intelligent control method is proposed. The basic controller gives a basic output value, predicts the trend of flue gas volume change in different stages of smelting through the neural network, and changes the working point of the system to adjust the fan speed in time. Fuzzy reasoning method is used to compensate the influence of bag filter on the system. The practical application shows that this method can obtain a good dust removal effect and reduce the operating cost of the dust removal system.