【摘要】 根据烧结矿化学成分与烧结工艺的预报、控制特点,采用了BP神经网络方法建立了烧结矿化学成分的预报模型。仿真实验的结果表明,模型具有较高的预测精度和较强的自学习功能,用拓扑结构为15-21-4的BP神经网络和0.6×10^-3的网络误差进行训练,模型的预报命中率在75%以上,充分验证了基于过程参数控制的烧结矿化学成分预测模型的准确性和有效性。更多还原

【Abstract】 According to the prediction and control characteristics of sintering process and sinter chemical composition,a prediction model of sinter chemical composition was established by using BP neural network.Simulation experimental results showed that the model has higher prediction precision and relatively strong self-learning ability.The predictive hit-ratio of random samples is over 75%by adopting BP neural network with the structure of 15-21-4 and network error of 0.6×10^-3,thus the accuracy and effectiveness of the quality prediction model based on process parameter control were verified.更多还原