Prediction of particulate matter in ambient environment in the surface coal mining areas: An analytical approach integrating Artificial Neural Network and Factor Analysis Mukhopadhyay Deboleena*, Dr. Sinha Suranjan** *Junior Research Fellow, BESU, Shibpur, DST sponsored project **Professor, Department of Mining Engineering, Bengal Engineering and Science University (BESU), Shibpur, Howrah-711103 Online published on 30 September, 2013. Abstract Air borne suspended particulate matter (SPM) and respirable particulate matter (PM10) are emitted from different mining operations in a surface mine. Several modeling techniques are used to predict dust concentration within mining areas. Dust propagation, in mining areas, is a complex phenomenon and relations between different meteorological and mine variables is non linear. Moreover, precise knowledge about generation of dust is not available, which is a prerequisite for any mathematical and statistical model. Modeling of atmospheric data where imprecise knowledgebase and data is available can be done using Artificial Neural Network (ANN). In this paper ANN modeling technique is used to build a model to predict suspended particulate matter and respirable dust particle concentration. To reduce dimensionality of the input space, due to presence of several variables in the model, Factor Analysis (FA) is used. The results improved after integration of ANN with FA. Top Keywords Suspended Particulate Matter (SPM), Particulate Matter of size below 10μ (PM10), Artificial Neural Network (ANN), Back Propagation Neural Network (BPNN), Factor Analysis (FA). Top |