中文摘要：

The concepts of neural network and tree structure are introduced to rare earth separations. The recursive relations and analytical expression for calculating the possible flow sheets and processes are derived in multi-component systems based on dichotomy. As an example, the application of the concepts is elucidated in detail in a separation of light rare earths containing La, Ce, Pr and Nd in the HEH(EHP)-HC1 system. The results show that this method is beneficial to summarizing and classifying the flow sheets and processes in rare earth separations, and is essential to the optimization of separation flow sheets. It can also be applied to the separation into group situations and other similar cases, too.

英文摘要：

The concepts of neural network and tree structure are introduced to rare earth separations. The recursive relations and analytical expression for calculating the possible flow sheets and processes are derived in multicomponent systems based on dichotomy. As an example, the application of the concepts is elucidated in detail in a separation of light rare earths containing La, Ce, Pr and Nd in the HEH(EHP)-HCl system. The results show that this method is beneficial to summarizing and classifying the flow sheets and processes in rare earth separations, and is essential to the optimization of separation flow sheets. It can also be applied to the separation into group situations and other similar cases, too.