关键词: 磨矿分级, 案例推理, 软测量, 推断控制

Abstract: Due to the difficulties of online measurement and direct closed-loop control of grinding particle in grinding-classification process, the case-based reasoning technique was adopted to build an online soft-sensor model for grinding particles. Then, based on this, an advanced inferential control scheme was proposed to achieve the closed-loop control of particle size. This advanced control system can automatically adjust the set-point of the lower-level PID controllers according to the operational conditions. As long as the outputs of the PID control loop can track their adjusted set-point, the closed-loop control of particle size will be achieved. The industrial application in a mineral processing plant showed that this proposed approach can achieve good performance of estimation and control of particle size.

Key words: Grinding-classification process, Case-based reasoning (CBR), Soft-sensor, Inferential control