APPLICATION OF NEURAL NETWORKS FOR THE PREDICTION OF ROCK FRAGMENTATION IN CHADORMALU IRON MINE

作者：Monjezi Masoud^*; Ahmadi Zabiholla; Khandelwal Manoj

来源：Archives of Mining Sciences, 2012, 57(3): 787-798.

DOI：10.2478/v10267-012-0051-0

• Fragmentation
• blasting operation
• ANN
• Chadormalu iron mine

Most open-pit mining operations employ blasting for primary breakage of the in-situ rock mass. Inappropriate blasting techniques can result in excessive damage to the wall rock, decreasing stability and increasing water influx. In addition, it will result in either over and/or under breakage of rocks. The presence of over broken rocks can result in decreased wall stability and require additional excavation. In contrast, the presence of under broken rocks may require secondary blasting and additional crushing. Since blasting is a major cost factor, both cases (under and over breakage) create additional costs reflected in the increase of the operation and maintenance of the machinery. Quick and accurate measurements of fragment size distribution are essential for managing fragmented rock and other materials. Various fragmentation measurement techniques are available and arc being used by industry/researchers but most of the methods arc time consuming and not precise. An ideally performed blasting operation enormously influences the overall mining cost. This aim can be achieved by proper prediction and attenuation of fragmentation. Prediction of fragmentation is essential for optimizing blasting operation. Poor performance of the empirical models for predicting fragmentation has urged the application of new approaches. In this paper, artificial neural network (ANN) method is implemented to develop a model to predict rock fragmentation size distribution due to blasting in Chadormalu iron mine, Iran. In the development of the proposed ANN model, ten parameters such as UCS, drilling rate, water content, burden, spacing, stemming, hole diameter, bench height, powder factor and charge per delay were incorporated. Training and testing of the model was performed by the back-propagation algorithm using 97 datasets. A four-layer ANN was found to be optimum with architecture of 10-7-5-1. A comparison has made between measured results of fragmentation with predicted results of fragmentation by ANN and multiple regression model. Sensitivity analysis was also performed to understand the effect of each influencing parameters on rock fragmentation.

• 出版日期2012