In order to solve measurement problem of copper liquid temperature so that qualified anode copper in the copper refining process can be obtained

some soft-sensing models of copper liquid temperature in the copper refining process including heat preservation

oxygenation and deoxidization were established based on the mechanism of heat transfer in the copper refining process

and some measures such as the pretreatment of the sampled data

the golden section method and parameter adjustment by using function link NN of the soft sensing model were introduced in the algorithm design of soft-sensing of copper liquid temperature. The application results revealed that the soft-sensing values of copper liquid temperature can describe real change of copper liquid temperature in the copper refining process. The soft-sensing model was useful for soft control of copper liquid temperature and for enhancing production efficiency and quality.