Performance Monitoring for Grate-kiln-cooler Process Based on Quality Prediction and Statistical Analysis

摘要

Grate-kiln-cooler is a predominant process of producing iron ore pellets in China. Since this process is of multi-variant, strong coupling and large hysteresis characteristics, a reliable performance monitoring is thus essential to ensure product quality and economic benefits. Aiming to efficiently monitor the pellet induration within this process, this study proposed an approach that combines statistical analysis and quality soft-measurement. Soft-measurement of pellet quality was accomplished based on the predicted time-temperature profile, and the soft-measured indices were on-line monitored using statistical control charts. Based on the proposed approach, a monitoring system for grate-kiln-cooler thermal process was developed for a domestic pelletizing plant. Running results demonstrate that our system can assist in the detection or diagnosis of abnormal thermal state and finally helps to stabilize the pellet production.