Abstract: In order to further optimize the control of buffer warehouse material weight in the process of high pressure roller mill，aiming at the process characteristics of inertia lag，complex working condition time-varying and strong nonlinearity in the process of material weight control，and the problem that most of the buffer warehouse material weight control is caused by manual on-site control，the labor intensity is large and the accuracy and timeliness of control are difficult to guarantee，an adaptive control method of high pressure roller mill weight based on 'end-edge-cloud' collaboration is studied. The real-time optimization setting method of roller material weight parameters based on case-based reasoning is studied，and the parameter controller is designed to realize the adaptive control of material weight parameters in the operation process，which solves the lag problem of traditional irregular and empirical control，and improves the intelligent level of concentrator.

Key words: case-based reasoning, end-edge-cloud, mineral processing, material weight