습식 비철금속 제련의 정액공정에서 불순물과 첨가제 간의 화학반응율 추정 방법 - 아연 제련을 중심으로 -

Abstract

The purification process removes impurity in the solution through a chemical reaction with an additive. The amount of additive depends on the estimate of its reaction rate with the impurity. We compare three estimation methods. First, a regression method formulates the complex chemical reaction between impurity and additive into a regression model. The reaction rate becomes the parameter of the model. Second, an adaptive method reflects the past errors detected by the first method. We develope a search procedure to find the optimal weights of the past errors in this method. Third, the Box-Jenkins method in time series models is applied. Through simulation these methods are analyzed and compared in various situations including a real case. It has been shown that the Box-Jenkins method is the most robust method to the situations. The case study determines the amount of zinc dust additive that reacts with the copper impurity in a hydrometallurgical zinc purification process.