Javier Taboada

FDM predictive methodology for subsidence due to flat and inclined coal seam mining

Abstract A subsidence estimation methodology to predict subsidence troughs due to flat and inclined coal seam exploitation is described. The foundations of the procedure are the correct definition of the rock mass behaviour model, its adequate characterization, and its implementation in the code “FLAC” (Fast Lagrangian Analysis of Continua was written by P.A. Cundall and is marketed by ITASCA Cons. Group Inc. of Minneapolis, MN, USA), which is based on the “Finite Difference Method” or FDM numerical modelling technique. The method is first validated for flat coal seam longwall mining with caving in British basins –because of the large number of data from this area– which includes subsidence measurements and characterization parameters. The method is then applied to gently inclined seams, and the results are seen to fit empirical observations. Finally, for steeply inclined and sub-vertical seams the proposed methodology allows observation of two different movement trends in the overlying strata; the former is perpendicular to the strata direction, in the same way as the one observed in flat seam subsidence; and the latter, which becomes the most important one in seams dipping by more than 75°, is parallel to the strata. These two different trends produce subsidence troughs presenting two relative subsidence maxima and extending along large surface areas. Thus, it would be unlikely to predict and assess these troughs, this may be a possible reason why so little coherent data about this kind of subsidence exists. These numerical observations of steeply inclined coal seams need to be validated empirically.

Explaining and predicting workplace accidents using data-mining techniques

Current research into workplace risk is mainly conducted using conventional descriptive statistics, which, however, fail to properly identify cause-effect relationships and are unable to construct models that could predict accidents. The authors of the present study modelled incidents and accidents in two companies in the mining and construction sectors in order to identify the most important causes of accidents and develop predictive models. Data-mining techniques (decision rules, Bayesian networks, support vector machines and classification trees) were used to model accident and incident data compiled from the mining and construction sectors and obtained in interviews conducted soon after an incident/accident occurred. The results were compared with those for a classical statistical techniques (logistic regression), revealing the superiority of decision rules, classification trees and Bayesian networks in predicting and identifying the factors underlying accidents/incidents.

Turbidity Prediction in a River Basin by Using Artificial Neural Networks: A Case Study in Northern Spain

Chemical and physical-chemical parameters define water quality and are involved in water body type and habitat determination. They support a biological community of a certain ecological status. Water quality controls involve a large number of measurements of variables and observations according to the European Water Framework Directive (Directive 2000/60/EC). In some cases, such as areas with especially critical uses or points in which potential pollution episodes are expected, the automatic monitoring is recommended. However, the chemical and physical-chemical measurements are costly and time consuming. Turbidity is shown as a key variable for the water quality control and it is also an integrative parameter. For this reason, the aim of this work is focused on this main parameter through the study of the influence of several water quality parameters on it. The artificial neural networks (ANNs) have been used in a wide range of biological problems with promising results. Bearing this in mind, turbidity values have been predicted here by using artificial neural networks (ANNs) from the remaining measured water quality parameters with success taking into account the synergistic interactions between the input variables in the Nalón river basin (Northern Spain). Finally, the main conclusions of this study are exposed.

A Conceptual Model for Analyzing the Risks Involved in the Transportation of Hazardous Goods: Implementation in a Geographic Information System

The transportation of hazardous goods by road implies a risk for both humans and the environment, in that an accident involving a vehicle transporting this kind of material may cause extensive material and environmental damage and might even endanger lives. For this reason, both public and private entities (e.g., insurance companies) have a growing interest in studies that assess the risks associated with hazardous goods transportation This article describes a method for calculating these risks. The risk is determined on the basis of a calculation of the probability of the occurrence of an accident and the gravity of the damage, which is in turn a function of the potential damage inherent in the goods being transported taken with the vulnerability of the environmental medium in which the accident takes place. The mathematical model proposed is easily implemented in a geographical information system that will produce risk maps delimiting the more potentially conflictive stretches of roadway.

Neural networks applied to discriminate botanical origin of honeys.

The aim of this work is develop a tool based on neural networks to predict the botanical origin of honeys using physical and chemical parameters. The managed database consists of 49 honey samples of 2 different classes: monofloral (almond, holm oak, sweet chestnut, eucalyptus, orange, rosemary, lavender, strawberry trees, thyme, heather, sunflower) and multifloral. The moisture content, electrical conductivity, water activity, ashes content, pH, free acidity, colorimetric coordinates in CIELAB space (L(∗), a(∗), b(∗)) and total phenols content of the honey samples were evaluated. Those properties were considered as input variables of the predictive model. The neural network is optimised through several tests with different numbers of neurons in the hidden layer and also with different input variables. The reduced error rates (5%) allow us to conclude that the botanical origin of honey can be reliably and quickly known from the colorimetric information and the electrical conductivity of honey.

Detection of outliers in water quality monitoring samples using functional data analysis in San Esteban estuary (Northern Spain).

Water quality controls involve large number of variables and observations, often subject to some outliers. An outlier is an observation that is numerically distant from the rest of the data or that appears to deviate markedly from other members of the sample in which it occurs. An interesting analysis is to find those observations that produce measurements that are different from the pattern established in the sample. Therefore, identification of atypical observations is an important concern in water quality monitoring and a difficult task because of the multivariate nature of water quality data. Our study provides a new method for detecting outliers in water quality monitoring parameters, using oxygen and turbidity as indicator variables. Until now, methods were based on considering the different parameters as a vector whose components were their concentration values. Our approach lies in considering water quality monitoring through time as curves instead of vectors, that is to say, the data set of the problem is considered as a time-dependent function and not as a set of discrete values in different time instants. The methodology, which is based on the concept of functional depth, was applied to the detection of outliers in water quality monitoring samples in San Esteban estuary. Results were discussed in terms of origin, causes, etc., and compared with those obtained using the conventional method based on vector comparison. Finally, the advantages of the functional method are exposed.

Functional classification of ornamental stone using machine learning techniques

Automated classification of granite slabs is a key aspect of the automation of processes in the granite transformation sector. This classification task is currently performed manually on the basis of the subjective opinions of an expert in regard to texture and colour. We describe a classification method based on machine learning techniques fed with spectral information for the rock, supplied in the form of discrete values captured by a suitably parameterized spectrophotometer. The machine learning techniques applied in our research take a functional perspective, with the spectral function smoothed in accordance with the data supplied by the spectrophotometer. On the basis of the results obtained, it can be concluded that the proposed method is suitable for automatically classifying ornamental rock.

Geostatistical study of the feldspar content and quality of a granite deposit

Abstract In order to characterise the saleable feldspar in a granite deposit, a methodology was developed in accordance with the exploitation process. This consisted of mechanically extracting the surface layer of the batholith and separating the feldspar from the quartz using the granulometric separation method, given that the size of the grains of the feldspar is greater than that of quartz. Following washing, grinding and magnetic separation of the feldspar in order to eliminate the ferromagnesium minerals, the saleable feldspar was characterised in terms of the factors that determine its market value, namely, its content in Al 2 O 3 , SiO 2 , Na 2 O and K 2 O. Following the opening of prospecting pits in the granite massif, samples were analysed in the laboratory using three different granulometric cuts and by reproducing the treatment process. The values for the quality variables of saleable feldspar were obtained, and the results were interpolated to the entire deposit using the kriging method. In order to summarise the information from the above-mentioned variables, a quality index was constructed using multivariate statistics and by employing market criteria, and subsequently, the values of the index were interpolated to the entire deposit using bidimensional kriging. The map of saleable quality feldspar from the deposit permits both affirmation of the treatment process yield for each granulometric cut and the planning of extraction from the deposit to obtain a homogeneous quality in the saleable feldspar.

Functional Pattern Recognition of 3D Laser Scanned Images of Wood-Pulp Chips

We evaluate the appropriateness of applying a functional rather than the typical vectorial approach to a pattern recognition problem. The problem to be resolved was to construct an online system for controlling wood-pulp chip granulometry quality for implementation in a wood-pulp factory. A functional linear model and a functional logistic model were used to classify the hourly empirical distributions of wood-chip thicknesses estimated on the basis of images produced by a 3D laser scanner. The results obtained using these functional techniques were compared to the results of their vectorial counterparts and support vector machines, whose input consisted of several statistics of the hourly empirical distribution. We conclude that the empirical distributions have sufficiently rich functional traits so as to permit the pattern recognition process to benefit from the functional representation.

Developing a new Bayesian Risk Index for risk evaluation of soil contamination

Industrial and agricultural activities heavily constrain soil quality. Potentially Toxic Elements (PTEs) are a threat to public health and the environment alike. In this regard, the identification of areas that require remediation is crucial. In the herein research a geochemical dataset (230 samples) comprising 14 elements (Cu, Pb, Zn, Ag, Ni, Mn, Fe, As, Cd, V, Cr, Ti, Al and S) was gathered throughout eight different zones distinguished by their main activity, namely, recreational, agriculture/livestock and heavy industry in the Avilés Estuary (North of Spain). Then a stratified systematic sampling method was used at short, medium, and long distances from each zone to obtain a representative picture of the total variability of the selected attributes. The information was then combined in four risk classes (Low, Moderate, High, Remediation) following reference values from several sediment quality guidelines (SQGs). A Bayesian analysis, inferred for each zone, allowed the characterization of PTEs correlations, the unsupervised learning network technique proving to be the best fit. Based on the Bayesian network structure obtained, Pb, As and Mn were selected as key contamination parameters. For these 3 elements, the conditional probability obtained was allocated to each observed point, and a simple, direct index (Bayesian Risk Index-BRI) was constructed as a linear rating of the pre-defined risk classes weighted by the previously obtained probability. Finally, the BRI underwent geostatistical modeling. One hundred Sequential Gaussian Simulations (SGS) were computed. The Mean Image and the Standard Deviation maps were obtained, allowing the definition of High/Low risk clusters (Local G clustering) and the computation of spatial uncertainty. High-risk clusters are mainly distributed within the area with the highest altitude (agriculture/livestock) showing an associated low spatial uncertainty, clearly indicating the need for remediation. Atmospheric emissions, mainly derived from the metallurgical industry, contribute to soil contamination by PTEs.