CAJ | 학술논문

利用基于物料衡算关系的机理模型、基于萃取平衡关系的半经验模型以及基于径向基函数-偏最小二乘法的萃余液的pH值预测模型,提出了一种钴湿法冶炼萃取过程质量指标软测量建模方法.机理模型用以描述过程的整体特性;半经验模型利用拟牛顿变尺度方法对其中未知参数进行估计,并用以逼近机理模型中的萃取平衡关系;基函数-偏最小二乘法则用来预测萃余液pH值.将所建立的软测量模型应用于实际钴湿法冶炼生产过程中,结果表明,该方法具有良好的预测精度.
이용기우물료형산관계적궤리모형、기우췌취평형관계적반경험모형이급기우경향기함수-편최소이승법적췌여액적pH치예측모형,제출료일충고습법야련췌취과정질량지표연측량건모방법.궤리모형용이묘술과정적정체특성;반경험모형이용의우돈변척도방법대기중미지삼수진행고계,병용이핍근궤리모형중적췌취평형관계;기함수-편최소이승법칙용래예측췌여액pH치.장소건립적연측량모형응용우실제고습법야련생산과정중,결과표명,해방법구유량호적예측정도.
A soft sensing modeling method for quality index of extraction process in cobalt hydrometallurgy is proposed. The soft sensing model is composed by first principal model based on material balance, semi experience model based on extraction equilibrium, and pH value of raffinate solution predict model based on radial basis function partial least squares (RBFPLS) First principle model describes the whole property of the process; semi experience model employs Quasi Newton variable metric (BFGS) method to identify the unknown parameters in it, for approaching the extraction equilibrium relationship in the first principal model; RBFPLS is utilized to predict the pH value of raffinate solution Applying the soft sensing model to practical industrial process of cobalt hydrometallurgy demonstrates that the method has satisfactory prediction precision.