Neural Networks In Mining Sciences – General Overview And Some Representative Examples
TL;DR: The article is intended to convince the readers that neural networks can be very useful also in mining sciences, and contains information how modern neural networks are built, how they operate and how one can use them.
Abstract: The many difficult problems that must now be addressed in mining sciences make us search for ever newer and more efficient computer tools that can be used to solve those problems. Among the numerous tools of this type, there are neural networks presented in this article – which, although not yet widely used in mining sciences, are certainly worth consideration. Neural networks are a technique which belongs to so called artificial intelligence, and originates from the attempts to model the structure and functioning of biological nervous systems. Initially constructed and tested exclusively out of scientific curiosity, as computer models of parts of the human brain, neural networks have become a surprisingly effective calculation tool in many areas: in technology, medicine, economics, and even social sciences. Unfortunately, they are relatively rarely used in mining sciences and mining technology. The article is intended to convince the readers that neural networks can be very useful also in mining sciences. It contains information how modern neural networks are built, how they operate and how one can use them. The preliminary discussion presented in this paper can help the reader gain an opinion whether this is a tool with handy properties, useful for him, and what it might come in useful for. Of course, the brief introduction to neural networks contained in this paper will not be enough for the readers who get convinced by the arguments contained here, and want to use neural networks. They will still need a considerable portion of detailed knowledge so that they can begin to independently create and build such networks, and use them in practice. However, an interested reader who decides to try out the capabilities of neural networks will also find here links to references that will allow him to start exploration of neural networks fast, and then work with this handy tool efficiently. This will be easy, because there are currently quite a few ready-made computer programs, easily available, which allow their user to quickly and effortlessly create artificial neural networks, run them, train and use in practice. The key issue is the question how to use these networks in mining sciences. The fact that this is possible and desirable is shown by convincing examples included in the second part of this study. From the very rich literature on the various applications of neural networks, we have selected several works that show how and what neural networks are used in the mining industry, and what has been achieved thanks to their use. The review of applications will continue in the next article, filed already for publication in the journal „Archives of Mining Sciences“. Only studying these two articles will provide sufficient knowledge for initial guidance in the area of issues under consideration here.
Citations
More filters
TL;DR: This paper discusses the automatic detection of metallic defects with a twofold procedure that accurately localizes and classifies defects appearing in input images captured from real industrial environments using a novel cascaded autoencoder (CASAE) architecture.
Abstract: Automatic metallic surface defect inspection has received increased attention in relation to the quality control of industrial products. Metallic defect detection is usually performed against complex industrial scenarios, presenting an interesting but challenging problem. Traditional methods are based on image processing or shallow machine learning techniques, but these can only detect defects under specific detection conditions, such as obvious defect contours with strong contrast and low noise, at certain scales, or under specific illumination conditions. This paper discusses the automatic detection of metallic defects with a twofold procedure that accurately localizes and classifies defects appearing in input images captured from real industrial environments. A novel cascaded autoencoder (CASAE) architecture is designed for segmenting and localizing defects. The cascading network transforms the input defect image into a pixel-wise prediction mask based on semantic segmentation. The defect regions of segmented results are classified into their specific classes via a compact convolutional neural network (CNN). Metallic defects under various conditions can be successfully detected using an industrial dataset. The experimental results demonstrate that this method meets the robustness and accuracy requirements for metallic defect detection. Meanwhile, it can also be extended to other detection applications.
288 citations
TL;DR: In this article, an early fault diagnostic technique based on acoustic signals was used for a single-phase induction motor, which can be also used for other types of rotating electric motors.
Abstract: An article describes an early fault diagnostic technique based on acoustic signals. The presented technique was used for a single-phase induction motor. The authors measured and analysed following states of the motor: healthy single-phase induction motor, single-phase induction motor with faulty bearing, single-phase induction motor with faulty bearing and shorted coils of auxiliary winding. A feature extraction method called MSAF-20-MULTIEXPANDED (Method of Selection of Amplitudes of Frequency – Multiexpanded) was discussed. The MSAF-20-MULTIEXPANDED was used to create feature vectors. The obtained vectors were classified by NN (Nearest Neighbour classifier), NM (Nearest Mean classifier) and GMM (Gaussian Mixture Models). The proposed technique can be used for diagnosis of the single-phase induction motors. It can be also used for other types of rotating electric motors.
286 citations
TL;DR: The authors develop an original method of the feature extraction of thermal images MoASoID (Method of Areas Selection of Image Differences), which compares many training sets together and it selects the areas with the biggest changes for the recognition process.
Abstract: Three-phase induction motors are used in the industry commonly for example woodworking machines, blowers, pumps, conveyors, elevators, compressors, mining industry, automotive industry, chemical industry and railway applications. Diagnosis of faults is essential for proper maintenance. Faults may damage a motor and damaged motors generate economic losses caused by breakdowns in production lines. In this paper the authors develop fault diagnostic techniques of the three-phase induction motor. The described techniques are based on the analysis of thermal images of three-phase induction motor. The authors analyse thermal images of 3 states of the three-phase induction motor: healthy three-phase induction motor, three-phase induction motor with 2 broken bars, three-phase induction motor with faulty ring of squirrel-cage. In this paper the authors develop an original method of the feature extraction of thermal images MoASoID (Method of Areas Selection of Image Differences). This method compares many training sets together and it selects the areas with the biggest changes for the recognition process. Feature vectors are obtained with the use of mentioned MoASoID and image histogram. Next 3 methods of classification are used: NN (the Nearest Neighbour classifier), K-means, BNN (the back-propagation neural network). The described fault diagnostic techniques are useful for protection of three-phase induction motor and other types of rotating electrical motors such as: DC motors, generators, synchronous motors.
182 citations
TL;DR: Support vector machine (SVM) with new 2level genetic optimizer (genetic training) and feature selection yielded the highest accuracy and F1-Score of 0.8849 and 0.8762 respectively in this work.
Abstract: Liver cancer is quite common type of cancer among individuals worldwide. Hepatocellular carcinoma (HCC) is the malignancy of liver cancer. It has high impact on individual’s life and investigating it early can decline the number of annual deaths. This study proposes a new machine learning approach to detect HCC using 165 patients. Ten well-known machine learning algorithms are employed. In the preprocessing step, the normalization approach is used. The genetic algorithm coupled with stratified 5-fold cross-validation method is applied twice, first for parameter optimization and then for feature selection. In this work, support vector machine (SVM) (type C-SVC) with new 2level genetic optimizer (genetic training) and feature selection yielded the highest accuracy and F1-Score of 0.8849 and 0.8762 respectively. Our proposed model can be used to test the performance with huge database and aid the clinicians.
81 citations
TL;DR: It is found that the performance of the proposed diagnostic methodology for the limited information is very effective for the same speed and load cases, and very encouraging for the intermediate speed case and the intermediate load case.
Abstract: This paper is focused on the development of a new SVM based fault diagnosis methodology for Induction Motors (IMs) in practical situation of limited data or information case. This work is of practical significance as the data is not always available at all operating conditions or in other word the data is limited for fault diagnosis. The vibration and current signals have been considered in this study since these signals are the most efficient to detect the electrical and mechanical faults as well as their severity in IMs. Ten different IM fault conditions, e.g. five electrical faults (i.e., the broken rotor bar, phase unbalance fault with two severity levels, and stator winding fault with two severity levels) and four mechanical faults (i.e., the bearing fault, bowed rotor, unbalanced rotor and misaligned rotor) with a healthy motor are considered. In order to develop the proposed diagnostic methodology, first the vibration and current data are acquired at various IM working conditions (i.e., the load and the speed) from an experimental setup. A number of fault features are then extracted using the raw time domain data. Further these features are fed into the SVM as inputs to diagnose IM faults. The fault diagnosis is first developed and checked for same speed cases, for the situation when the data is available at required loads and speeds. Diagnosis is then extended for an intermediate speed case and an intermediate load case for taking care of the situations, where the required information or data is not readily available at required speeds and loads. The aim of this study is to check the prediction performance of the proposed diagnostic methodology for the limited information. This study is very significant for the practical point of view since symptoms database are not available for all the cases or sometimes difficult to obtain at all IM working conditions. To investigate the robustness of the present diagnostic methodology, it is checked at several working conditions. It is found that the performance is very effective for the same speed and load cases, and very encouraging for the intermediate speed case and the intermediate load case.
49 citations
References
More filters
Book•
14 Aug 2014
TL;DR: Exploring Neural Networks with C# presents the important properties of neural networks while keeping the complex mathematics to a minimum and presents complicated information about neural networks structure, functioning, and learning in a manner that is easy to understand.
Abstract: The utility of artificial neural network models lies in the fact that they can be used to infer functions from observationsmaking them especially useful in applications where the complexity of data or tasks makes the design of such functions by hand impractical.Exploring Neural Networks with C# presents the important properties of neural networkswhile keeping the complex mathematics to a minimum. Explaining how to build and use neural networks, it presents complicated information about neural networks structure, functioning, and learning in a manner that is easy to understand.Taking a "learn by doing" approach, the book is filled with illustrations to guide you through the mystery of neural networks. Examples of experiments are provided in the text to encourage individual research. Online access to C# programs is also provided to help you discover the properties of neural networks.Following the procedures and using the programs included with the book will allow you to learn how to work with neural networks and evaluate your progress. You can download the programs as both executable applications and C# source code from http://home.agh.edu.pl/~tad//index.php?page=programy&lang=en
62 citations
TL;DR: In this article, a novel high-pressure apparatus with various abilities in hydrate investigation fields has been designed, constructed and fully described in the present paper, which can be used in future inhibitor dosing devices.
Abstract: A novel high-pressure apparatus with various abilities in hydrate investigation fields has been designed, constructed and fully described in the present paper. In order to achieve an appropriate understanding of the gas hydrate behavior in formation and destabilization, series of laboratory experiments with six different gas mixtures were done and more than 130 hydrate equilibrium points in the pressure range of about 450–3000 psia were recorded. Different methods of hydrate formation prediction were discussed and finally the new promising neural networks method was used. Because of the previous works defects in accurate hydrate formation prediction via neural networks, a new use of neural networks was introduced. Testing and validation of the new neural networks method indicates that it is a reliable technique for the accurate prediction of hydrate formation conditions for generalized gas systems and can be used in future automatic inhibitor dosing devices.
60 citations
TL;DR: The proposed intelligent model can be considered as an alternative model to predict the efficiency of chemical flooding in oil reservoir when the required experimental data are not available or accessible.
Abstract: Application of chemical flooding in petroleum reservoirs turns into hot topic of the recent researches. Development strategies of the aforementioned technique are more robust and precise when we consider both economical points of view (net present value, NPV) and technical points of view (recovery factor, RF). In current study many attempts have been made to propose predictive model for estimation of efficiency of chemical flooding in oil reservoirs. To gain this end, a couple of swarm intelligence and artificial neural network (ANN) is employed. Also, lucrative and high precise chemical flooding data banks reported in previous attentions are utilized to test and validate proposed intelligent model. According to the mean square error (MSE), correlation coefficient, and average absolute relative deviation, the suggested swarm approach has acceptable reliability, integrity and robustness. Thus, the proposed intelligent model can be considered as an alternative model to predict the efficiency of chemical flooding in oil reservoir when the required experimental data are not available or accessible.
60 citations
TL;DR: Two intelligent approaches including multilayer perceptron (MLP) neural network and least squares support vector machine (LSSVM) algorithm are utilized to predict optimum stripping gas flow rate in natural gas dehydration systems.
Abstract: Natural gas is an extremely vital supply of energy. Compared to other fossil fuels, demand for natural gas is experiencing a considerable growth due to its accessibility, availability, versatility and smaller environmental footprint. Glycol dehydration is the most common and economic method of water removal from natural gas streams. The water content of the dehydrated gas depends primarily on the lean triethylene glycol (TEG) concentration. Injecting stripping gas into the reboiler is taken into account as an effective way to improve the glycol concentration. In this article, two intelligent approaches including multilayer perceptron (MLP) neural network and least squares support vector machine (LSSVM) algorithm are utilized to predict optimum stripping gas flow rate in natural gas dehydration systems. Furthermore, a simple mathematical tool is presented for the application of interest. Based on the statistical analysis, an excellent match is noticed between the values obtained from the predictive tools (e.g., MLP, LSSVM and the empirical equations) and the real data so that the average absolute relative deviation percent (AARD %) is determined to be lower than 0.01%.
56 citations
TL;DR: In this article, a back-propagation neural network (ANN) was used for petrographic classification of carbonate-siliciciclastic rocks using elastic, mineralogical, and textural information.
Abstract: Petrographic class identification is of great importance to petroleum reservoir characterization and wellbore economic viability analysis, and is usually performed using core or geophysical log analysis. The coring process is costly, and well log analysis requires highly specific knowledge. Thus, great interest has arisen in new methods for predicting the lithological and textural properties of a wide area from a small number of samples. The artificial neural network (ANN) is a computational method based on human brain function and is efficient in recognizing previously trained patterns. This paper demonstrates petrographic classification of carbonate-siliciclastic rocks using a back-propagation neural network algorithm supported by elastic, mineralogical, and textural information from a well data set located in the South Provence Basin, in the southwest of France. The accuracy of the testing suggests that an ANN application offers an auxiliary tool for petrographic classification based on well data, specifically for prediction intervals in wells that have not been sampled or wells adjacent to sampled wells.
45 citations