Investigation of general regression neural network architecture for grade estimation of an Indian iron ore deposit
- PDF / 3,962,981 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 52 Downloads / 208 Views
ORIGINAL PAPER
Investigation of general regression neural network architecture for grade estimation of an Indian iron ore deposit Agam Das Goswami 1 & M. K. Mishra 1 & Dipti Patra 2
Received: 16 August 2016 / Accepted: 23 January 2017 # Saudi Society for Geosciences 2017
Abstract The mineral resource estimation requires accurate prediction of the grade at location from limited borehole information. It plays the dominant role in the decision-making process for investment and development of various mining projects and hence become an important and crucial stage. This paper evaluvates the use of two distinct artificial neural network (ANN)-based models, general regression neural network (GRNN) and multilayer perceptron neural network (MLP NN), to improve the grade estimation from Koira iron ore region in Sundargarh district, Odisha. ANN-based models capture the inherent complex structure of mineral deposits and provide a reliable generalization of the iron grade. The ANNbased approach does not require any preliminary geological study and is free from any statistical assumption on the raw data before its application. The GRNN is a one-pass learning algorithm and does not require any iterative procedure for training less complex structure and requires only one learning parameter for optimization. In this investigation, the spatial coordinates and multiple lithological units were taken as input variables and the iron grade was taken as the output variable. The comparative analysis of these models has been carried out and the results obtained were validated with traditional geostatistical method ordinary kriging (OK). The GRNN model outperforms the other methods, i.e. MLP and OK, with respect to generalization and predictability of the grades at an un-sampled location.
* Agam Das Goswami [email protected]
1
Department of Mining Engineering, NIT Rourkela, Rourkela 769008, India
2
Department of Electrical Engineering, NIT Rourkela, Rourkela 769008, India
Keywords Grade estimation . General regression neural network (GRNN) . Multilayer perceptron neural network (MLP NN) . Ordinary kriging (OK)
Introduction Mineral resource estimation is one of the vital stages and the most complicated aspects of mining. It plays major role in the decision-making process for the investment in mining, pit designing, production scheduling and grade control as discussed elsewhere (David 1977; Goovaerts 1997; Ke 2002; Mahmoudabadi et al. 2009; Tahmasebi and Hezarkhani 2010, 2012). The major goal of mineral resource estimation is to assess the mineral grade at location with limited borehole information (Li et al. 2013). The spatial distribution of mineral deposit is random in nature due to complex geological activities over the long period of time. Therefore, mineral resource estimation at an un-sampled location is challenging (Dutta et al. 2010). The traditional approaches such as inverse distance weighing; conventional geostatistical methods such as kriging and its various versions (simple kriging, ordinary kriging (OK), lognor
Data Loading...
