基于多相流超图神经网络的烧结矿质量预报评价系统

A Sintering Quality Prediction and Evaluation System Based on Multiphase Flow Supergraph Neural Network

“双碳”背景下,钢铁作为能耗重点行业,配矿结构亟待优化。为顺应时代洪流,本项目提出基于多相流超图神经网络的烧结矿质量预报评价系统。首先,收集烧结矿混合料化学成分、内返率等指标值。其次,使用超图学习文本数据。最后,建立烧结矿质量预报评价系统。该系统可应用于优化配矿,使烧结原料具有良好的制粒性能和成矿性能,实现高产、优质、低耗烧结生产。

Under the background of"dual carbon",steel,as a key energy consuming industry,urgently needs to optimize its ore blending structure.To keep up with the tide of the times,this project proposes a sintering ore quality prediction and evaluation system based on multiphase flow hypergraph neural network.Firstly,collect the chemical composition,internal return rate,and other indicator values of the sintered ore mixture.Secondly,use hypergraphs to learn text data.Finally,establish a quality prediction and evaluation system for sintered ore.This system can be applied to optimize ore blending,enabling sintering raw ma-terials to have good granulation and mineralization properties,achieving high yield,high quality,and low consumption sinter-ing production.