An Optimization-Based Clustering Method used for an Efficient Oil and Gas Fields Pipelines Design

An oil and gas pipelines network design is essential for the field production engineer. Therefore a well-studied plan of the area where the pipeline network must be installed. In this work, we attempt to solve the optimal classification of the well-heads zones using the clustering methodology. The proposed method aims to minimize the lengths of the required pipelines within any particular zone. Using the metaheuristic methods has been approved through the years. The first step involves detecting the well-head locations required to be divided into non-intersecting regions. Later, the well-head distances from their central points (manifolds) are calculated iteratively by the hybridized harmony search method. The suggested approach combines the hybrid harmony search method with the traditional unsupervised learning method. The internal measurements of each region's coherency are utilized to measure the compactness of each area. Utah FORGE Seismic Stations and Wells GPS Survey Data is used. The clustering accuracy obtained exceeded 90% percent using the F-measure score.