Snehamoy Chatterjee

Reliability estimation using a genetic algorithm-based artificial neural network: An application to a load-haul-dump machine

In this study, a neural network-based model for forecasting reliability was developed. A genetic algorithm was applied for selecting neural network parameters like learning rate (@h) and momentum (@m). The input variables of the neural network model were selected by maximizing the mean entropy value. The developed model was validated by applying two benchmark data sets. A comparative study reveals that the proposed method performs better than existing methods on benchmark data sets. A case study was conducted on a load-haul-dump (LHD) machine operated at a coal mine in Alaska, USA. Past time-to-failure data for the LHD machine were collected, and cumulative time-to-failure was calculated for reliability modeling. The results demonstrate that the developed model performs well with high accuracy (R^2=0.94) in the failure prediction of a LHD machine.

Genetic algorithms for feature selection of image analysis-based quality monitoring model: An application to an iron mine

Measuring the quality parameters of materials at mines is difficult and a costly job. In this paper, an image analysis-based method is proposed efficiently and cost effectively that determines the quality parameters of material. The image features are extracted from the samples collected from a mine and modeled using neural networks against the actual grade values of the samples generated by chemical analysis. The dimensions of the image features are reduced by applying the genetic algorithm. The results showed that only 39 features out of 189 features are sufficient to model the quality parameter. The model was tested with the testing data set and the result revealed that the estimated grade values are in good agreement with the real grade values (R^2=0.77). The developed method was then applied to a case study mine of iron ore. The case study results show that proposed image-based algorithm can be a good alternative for estimating quality parameters of materials at a mine site. The effectiveness of the proposed method was verified by applying it on a limestone deposit and the results revealed that the method performed equally well for the limestone deposit.

Dimensional Reduction of Pattern-Based Simulation Using Wavelet Analysis

A pattern-based simulation technique using wavelet analysis is proposed for the simulation (wavesim) of categorical and continuous variables. Patterns are extracted by scanning a training image with a template and then storing them in a pattern database. The dimension reduction of patterns in the pattern database is performed by wavelet decomposition at certain scale and the approximate sub-band is used for pattern database classification. The pattern database classification is performed by the k-means clustering algorithm and classes are represented by a class prototype. For the simulation of categorical variables, the conditional cumulative density function (ccdf) for each class is generated based on the frequency of the individual categories at the central node of the template. During the simulation process, the similarity of the conditioning data event with the class prototypes is measured using the L2-norm. When simulating categorical variables, the ccdf of the best matched class is used to draw a pattern from a class. When continuous variables are simulated, a random pattern is drawn from the best matched class. Several examples of conditional and unconditional simulation with two- and three- dimensional data sets show that the spatial continuity of geometric features and shapes is well reproduced. A comparative study with the filtersim algorithm shows that the wavesim performs better than filtersim in all examples. A full-field case study at the Olympic Dam base metals deposit, South Australia, simulates the lithological rock-type units as categorical variables. Results show that the proportions of various rock-type units in the hard data are well reproduced when similar to those in the training image; when rock-type proportions between the training image and hard data differ, the results show a compromise between the two.

Determinants of work injuries in mines – an application of structural equation modelling

In spite of stringent regulations and much attention towards reducing risks in the physical environment, the mining industry continues to be associated with high levels of accidents, injuries and illnesses. Only engineering solutions to accident prevention are inappropriate unless coupled with focused attention to the attitudes and behaviours of the mineworkers in coping with the inherent physical, technical and situational risks. The present study identified these various risk factors and analysed their influences on work injury in a causal framework. Data were collected from an underground coalmine of India. The pattern and strength of relationships of 16 causal factors with work injuries were assessed through structural equation modelling. The case study results showed that negatively personified individuals are of major concern for safety improvement in the mine studied. They not only fail to avoid work injuries, they are unable to extend safe work behaviours in their work. The variable safety environment is negatively affected by personality, whereas social support has a positive relationship with safety environment. The variable job hazards appeared to have a significant relationship with job involvement, which has a negative relationship with work injury. Elimination of negative behaviours must be focused and committed by the mine safety management. Long term planning through (i) identification of negative individuals, (ii) proper councelling of adverse effects of negative behaviours, and (iii) special training with psychological treatment is highly required. Identification may begin while recruiting new workers through interview. Proper allocation of jobs (right person for right job) may be a judicial solution to this end.

Multi-scale stochastic simulation with a wavelet-based approach

Conditional simulation of random fields based on integrating multi-scale spatial data using wavelets is detailed herein. Course scale data are first simulated using sequential Gaussian co-simulation. For all finer scales, wavelet coefficients are simulated by using a template matching algorithm borrowing information from a training image. Spatial up-scaling is performed through the inverse wavelet transformation. Examples using an exhaustive dataset show that the proposed method works well and is robust when changing the amount of hard data and resolution of secondary data. Sensitivity analysis shows that the selection of a suitable template size can improve the performance of the proposed method. A comparative study shows that the proposed algorithm is computationally faster than the well-known simpat and filtersim algorithms, as well as reasonably accurate in terms of reproduction of continuity in complex geologic environments.

Ore grade estimation of a limestone deposit in India using an Artificial Neural Network

This study describes a method used to improve ore grade estimation in a limestone deposit in India. Ore grade estimation for the limestone deposit was complicated by the complex lithological structure of the deposit. The erratic nature of the deposit and the unavailability of adequate samples for each of the lithogical units made standard geostatistical methods of capturing the spatial variation of the deposit inadequate. This paper describes an attempt to improve the ore grade estimation through the use of a feed forward neural network (NN) model. The NN model incorporated the spatial location as well as the lithological information for modeling of the ore body. The network was made up of three layers: an input, an output and a hidden layer. The input layer consisted of three spatial coordinates (x, y and z) and nine lithotypes. The output layer comprised all the grade attributes of limestone ore including silica (SiO2), alumina (Al2O3), calcium oxide (CaO) and ferrous oxide (Fe2O3). To justify the use of the NN in the deposit, a comparative evaluation between the NN method and the ordinary kriging was performed. This evaluation demonstrated that the NN model decisively outperformed the kriging model. After the superiority of the NN model had been established, it was used to predict the grades at an unknown grid location. Prior to constructing the grade maps, lithological maps of the deposit at the unknown grid were prepared. These lithological maps were generated using indicator kriging. The authors conclude by suggesting that the method described in this paper could be used for grade-control planning in ore deposits.

CDFSIM: Efficient Stochastic Simulation Through Decomposition of Cumulative Distribution Functions of Transformed Spatial Patterns

Simulation of categorical and continuous variables is performed using a new pattern-based simulation method founded upon coding spatial patterns in one dimension. The method consists of, first, using a spatial template to extract information in the form of patterns from a training image. Patterns are grouped into a pattern database and, then, mapped to one dimension. Cumulative distribution functions of the one-dimensional patterns are built. Patterns are then classified by decomposing the cumulative distribution functions, and calculating class or cluster prototypes. During the simulation process, a conditioning data event is compared to the class prototype, and a pattern is randomly drawn from the best matched class. Several examples are presented so as to assess the performance of the proposed method, including conditional and unconditional simulations of categorical and continuous data sets. Results show that the proposed method is efficient and very well performing in both two and three dimensions. Comparison of the proposed method to the filtersim algorithm suggests that it is better at reproducing the multi-point configurations and main characteristics of the reference images, while less sensitive to the number of classes and spatial templates used in the simulations.

Ensemble Support Vector Machine Algorithm for Reliability Estimation of a Mining Machine

In this study, a support vector machine (SVM)-based ensemble model was developed for reliability forecasting. The hyperparameters of the SVM were selected by applying a genetic algorithm. Input variables of the SVM model were selected by maximizing the mean entropy value. The diverse members of the ensemble model were obtained by a k-means clustering algorithm, and one ensemble member was selected from each cluster by choosing the closest from the cluster center. The optimum cluster number was selected using the Davies–Bouldin index. The developed model was validated by a benchmark turbocharger data set. A comparative study reveals that the proposed method performs better than existing methods on benchmark data sets. A case study was conducted investigating a dumper operated at a coal mine in India. Time-to-failure historical data for the dumper were collected, and cumulative time to failure was calculated for reliability forecasting. Study results demonstrate that the developed model performs well with high accuracy (R2 = 0.97) in the prediction of dumper failure, and a comparison with other methods demonstrates the superiority of the proposed ensemble SVM model. Copyright

Production phase and ultimate pit limit design under commodity price uncertainty

Open pit mine design optimization under uncertainty is one of the most critical and challenging tasks in the mine planning process. This paper describes the implementation of a minimum cut network flow algorithm for the optimal production phase and ultimate pit limit design under commodity price or market uncertainty. A new smoothing splines algorithm with sequential Gaussian simulation generates multiple commodity price scenarios, and a computationally efficient stochastic framework accommodates the joint representation and processing of the mining block economic values that result from these commodity price scenarios. A case study at an existing iron mining operation demonstrates the performance of the proposed method, and a comparison with conventional deterministic approach shows a higher cumulative metal production coupled with a 48% increase in the net present value (NPV) of the operation.

Rock-type classification of an iron ore deposit using digital image analysis technique

In this paper, the rock types of an iron ore deposit were classified using the digital image analysis technique. The image acquisition and analysis of blasted rocks were conducted in a laboratory for six different rock types. A total of 189 features were extracted from the individual rock samples using best-suited segmentation technique selected by validation study. The neural network technique was applied for rock classification model using image features. Five principal components, which accounts for 95% of total data variance, were selected as input parameters for the model. The misclassification error of the model for testing data was 2.4%.