Prediction of Clean Coal Yield and Ash at Coal Washeries using a Combination of Neural Network and Empirical Models

摘要

Prediction of clean coal yield from coal washeries has always been of utmost importance to washery personnel. The prediction, through models, is cumbersome which involves conducting a large number of sink float tests and drawing of graphs to determine efficiency of separation units (Ep). Though the principles of dense medium cyclone separation and froth flotation are well known, not much is known about the actual separation process taking place inside these cyclones or during froth flotation. The processes are more of a 'black box'. For such situations, artificial neural network (ANN) has been found to be the most suitable tool, which can give reliable prediction from available data. ANN is an artificial intelligence tool that can be used for determining co-relations in a multi-variable non-linear system for extracting information from raw data. The paper describes in detail, the use of ANN for predicting the performance of froth flotation circuit and the use of empirical model and ANN for predicting the performance of DM cyclones.