【摘要】 针对稀土萃取生产现场采集到的有效建模样本数据少，易导致模型预测精度不高等小样本问题，提出采用随机配置网络(SCN)生成虚拟样本进行组分含量预测的方法。以真实样本确定的SCN模型为依据，根据隐含层与输出层、输入层与隐含层的之间的映射关系，采用中点插值方法生成虚拟样本；然后混合原始真实样本与虚拟样本，建立基于SCN的铈镨/钕(CePr/Nd)组分含量预测模型。通过稀土萃取现场数据验证，结果表明：本文方法可以实现稀土萃取过程现场组分含量的快速、准确检测。更多还原

【Abstract】 Aiming at the small sample problems of less data of effective modeling sample, acquired in rare earth extraction production site, resulting in low model prediction precision, a method for predicting the component content of rare earth extraction process by using stochastic configuration network(SCN)to generate virtual sample is proposed.The virtual sample is generated by the midpoint interpolation method in conjunction with the mapping relationship between the hidden layer and the output layer, the input layer and the hidden layer of the SCN model based on the real sample.The SCN-based prediction model of cerium praseodymium/neodymium(CePr/Nd)component content is established by mixing the original real sample and the virtual sample.The model is evaluated by the data collected on the CePr/Nd extraction field.The experimental results show that the proposed method is suitable for rapid and accurate prediction of the component content of the rare earth extraction process.更多还原