基于遗传神经网络的氧化铝浓度预测

Alumina Concentration Prediction Based on Genetic Neutral Network

为了精确地对铝电解过程中氧化铝浓度进行控制,通过分析槽电阻与氧化铝浓度的关系,结合铝电解工艺的特点构建基于BP神经网络的氧化铝浓度预测模型。同时,提出利用遗传算法来优化BP神经网络的结构和网络参数,然后利用优化后的网络进行氧化铝浓度进行预测。仿真结果显示,该模型能很好地对氧化铝浓度进行预测,且误差较小。利用遗传神经网络对氧化铝浓度进行预测快速、有效,对实现铝电解过程氧化铝精确控制有着重要意义。

To accurately control the alumina concentration during the process of aluminum electrolysis,an alumina concentration prediction model based on BP neural network was established in combination with the characteristics of the aluminum electrolysis process and analysis of the relation between cell resistance and alumina concentration.It is proposed that the structure and network parameters of the BP neural network are optimized with the genetic algorithm,and the alumina concentration prediction is performed with the optimized network.The simulation results indicate that it is effective to apply the GA-BP network to predict the alumina concentration,and the prediction error is very small.The prediction method mentioned above is fast and effective.It has great significance to the accurate control of the alumina concentration in the process of aluminum electrolysis.