End point prediction of basic oxygen furnace (BOF) steelmaking based on improved bat-neural network

Abstract

A mixed bat optimization algorithm based on chaos and differential evolution (CDEBA) is proposed for the endblow process of basic oxygen furnance (BOF) after sub-lance detection, and a prediction model based on BP neural network optimized by chaotic differential bat algorithm (CDEBA-NN) is presented. The simulation results show that the prediction model of carbon content achieves a hit rate of 94 % with the error range of 0,005 %, and 90 % for temperature with the error range of 15 °C, the accuracy is higher than the traditional neural network model, and then it verifies the effectiveness of the proposed model.

Keywords

• steelmaking
• BOF
• end point prediction
• back propagation (BP) neural network
• bat algorithm