02-09-2017, 11:35 AM
Cutting analysis is an important and crucial task to detect and prevent problems during the drilling of oil wells. Several studies have been developed for drilling inspection, but none of them are concerned with analyzing the cut generated in vibrating slate shakers. Here we proposed a system to analyze the concentration of the cut in vibrating shakers of slate, which can indicate problems during the drilling process of oil wells, so that the collapse of the walls of the well. Cutting images are acquired and sent to the data analysis module, whose main objective is to extract features and classify the frames according to one of the three previous classes of cut volume. A collection of supervised classifiers was applied to allow comparisons of their accuracy and efficiency. We used the Optimal Path Forest (OPF), Artificial Neural Network using Multilayer Perceptron (ANN-MLP), Support Vector Machines (SVM) and a Bayesian Classifier (BC) for this task. The first one outperformed all other classifiers. Recall that we are also the first to introduce the OPF classifier in this field of knowledge. Very good results demonstrate the robustness of the proposed system, which can also be integrated with another common system (Mud-Logging) to improve the efficiency of the latter.