基于RF-PCCs和RBF的焦炭质量预测模型研究

Study on Prediction Model for Coke Quality Based on RF-PCCs and RBF

在炼焦生产实践中,焦炭质量预测模型的准确性对降低配煤成本和优质炼焦煤的消耗具有重要作用。为此,基于径向基函数神经网络模型,通过随机森林和皮尔逊相关系数算法共同进行特征选择,建立了焦炭质量预测模型,并以某钢厂炼焦部生产焦炭的各项实际数据进行仿真。结果表明,该模型预测精度高,泛化能力强,可为炼焦生产提供一定的理论依据。

In the practice of coking production,the accuracy in terms of the prediction model for coke quality plays an important role in reducing the cost of coal blending and the consumption of high-quality coking coal.Therefore,based on radial basis function neural network model,the prediction model for coke quality was established through carrying out common feature selections by random forest and Pearson correlation coefficient algorithm.After that the actual data on coke-making from coking department of a steel plant in operation were simulated depending on the model.The simulation results showed that the model had high prediction accuracy and strong generalization ability,which could provide a certain theoretical basis for coking production.