Search

Studies conducted by several researchers suggest that a large variance exists in the morphological integrity of rocks when subjected to thermal treatment. The extent of thermal damage, D(T), can be quantified by analyzing the change in either the elastic modulus, ultrasonic wave velocities or the acoustic emission signals. However, these require the use of sophisticated laboratory equipment, which may not be readily available. Additionally, the shape and size of the sample has to adhere to the specifications that have been mandated for the corresponding experiments. This would further introduce delay in the process of assessing the damage. Therefore, in this study, new predictive models have been developed, which can predict the extent of damage (D(T)) from the physical properties of thermally-modified fine-grained Dholpur sandstone. The sandstone is a primary construction material, and has been widely used in several Indian monuments of historic and political importance. The models have been developed using statistical and soft-computing tools such as multivariate regression analysis (MVRA), artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS). The physical properties viz. temperature (T), density (D), porosity (P), thermal expansion coefficients (E-L and E-v) and ultrasonic wave velocities (V-P and V-S), serve as predictor variables. The efficacy of the models has been tested by calculating the performance indices, namely, coefficient of determination (R2), mean absolute percentage error (MAPE), root mean square error (RMSE) and variance account for (VAF). The results suggest that the ANFIS model has the best prediction capacity.

This paper deals with the application of a support vector machine (SVM) optimization technique to predict the blast-induced ground vibration. Peak particle velocity (PPV) is an important parameter to be kept under control to minimize the damage caused due to the ground vibration. A number of previous researchers have tried to use different empirical methods to predict PPV, but these empirical equations have their limitations due to its less versatile application and acceptability from field conditions. Therefore, it is difficult to apply these empirical equations to predict PPV because they are based on limited parameters which does not really reflect and connect with real influencing parameters. In this paper, SVM technique is used for the prediction of PPV by incorporating blast design and explosive parameters, and the suitability of one technique over other has been tested based on the results. To avoid the biasness in man-made choice of parameters of SVM, we have used the chaos optimization algorithm to find the optimal parameters which can help the model to enhance the learning efficiency and capability of prediction. Datasets have been obtained from one of the large opencast mine from southeastern coalfield limited, Chhattisgarh, India. One hundred and twenty-seven datasets were used to establish SVM architecture, and 10 datasets have been randomly chosen for validation of SVM model to see its prediction potential. The results obtained have been compared with different vibration predictors, multivariate regression analysis, artificial neural network and the superiority of application on SVM over previous methodology. The mean absolute percentage error using SVM is very low (0.001) as compared to other predictors indicate its better prediction capability.

An attempt has been made to evaluate and predict the blast-induced ground vibration and frequency by incorporating rock properties, blast design and explosive parameters using the artificial neural network (ANN) technique. A three-layer ,feed-forward back-propagation neural network having 15 hidden neurons, 10 input parameters and two output parameters were trained using 154 experimental and monitored blast records from one of the major producing surface coal mines in India. Twenty new blast data sets were used for the validation and comparison of the peak particle velocity (PPV) and frequency by ANN and other predictors. To develop more confidence in the proposed method, same data sets have also been used for the prediction of PPV by commonly used vibration predictors as well as by multivariate regression analysis (MVRA). Results were compared based on correlation and mean absolute error (MAE) between monitored and predicted values of PPV and frequency. (C) 2009