Research On The Prediction Of Operating Condition Of Blast Furnace Blower Based On Neural Network

The blast furnace blower in ironmaking process plays an important role as one of the key equipment of iron, Due to the continuity of the blast furnace smelting blower,it requirements evenly to supply a certain amount of air and a certain air pressure to overcome the resistance of the air supply system and stock column and make the blast furnace top pressure is constant. In the whole process of smelting, due to the change of raw materials, fuel, and operating conditions, it often causes furnace condition change.In order to reduce the accident frequency,save maintenance costs, reduce downtime and improve economic efficiency, it is necessary to predict operational status of the blast furnace blower.For this kind of situation, in order to achieve the prediction of blast furnace blower operation condition, First of all, according to the relationship between the technological process of each link,on the basis of accumulated a large amount of data information, collected field data is applied to the hard and soft threshold value of wavelet transform to compromise algorithm to process the wavelet coefficients and filter out noise. Then prediction model is established using BP neural network to adjust and train the structure of network, and the network output value is applied in the inverse transform for actual forecast. Finally, it implements the bearing shell temperature prediction before the blast furnace blower, and obtain a good prediction results, by Matlab simulation. Results show that: In the premise using wavelet transform to effectively filter the noise in the data, the established BP neural network can effectively realize the prediction of the blast furnace blower operation.Prediction system of the blast furnace blower operation using the Visual Basic development can be used as a platform for BP neural network model training and the prediction for the blast furnace blower performance.