Designing of an intelligent system using acoustic signatures to predict the particle size of a tumbling mill in real time environment

Abstract

Increasing emphasis on mineral productivity by reducing the wastage of newlinematerial and to control the quality of the product, an impetus has been newlineconcentrated to the research on the different methodologies in the field newlineof mineral processing system. In mining industry different newlinemethodologies are studied for diagnosis, monitor, control, model and newlineoptimise the grinding procedure of the ores used for different purposes. newlineOne of the major problems in this mining processing is the raw material newlinewastage and to get rid of it. One of the biggest challenges facing by the newlineresearchers is to place a raw sensor to retrieve some meaningful data newlineduring the crushing operation so that the process can be automated. newlineThe grinding of the raw materials with in a tumbling mill like ball mill, newlineis continued for a certain time instances and then the mill is opened and newlinechecked the size of the particles. From the crushed materials, the desired newlineparticle size ranges are collected and rest of the materials whose size newlineranges are larger are fed into the mill for further crushing. But the under newlinesized materials are the wastage which are nothing but the high valued newlineminerals. Due to huge dust and the presence of grinding media no sensors newlinecan be fitted inside the mill to acquire knowledge using which the system newlinecan be stopped after achieving the desired particle size distribution newline