Optimizing the design of pipelines for drinking water systems using deep neural networks in the Huancabamba District

This dataset corresponds to the study titled Optimizing the design of pipelines for drinking water systems using deep neural networks in the Huancabamba District. The data were collected from 75 rural water supply projects located in the Huancabamba District, Piura Region, Peru. Each project includes hydraulic, topographic, and design parameters such as flow rate, pipe diameter, velocity, pressure, and elevation profiles, which were used to train and validate a deep neural network model developed in Python and TensorFlow. The dataset includes structured files in CSV and Excel formats containing: General project data: geographic coordinates of intake and reservoir points, elevations, and pipe material information. Topographic profiles: surveyed elevation points along the transmission lines. Hydraulic design parameters: flow, pressure, velocity, and diameter for each section of the pipeline. Predicted outputs: optimized pipeline alignment and design parameters generated by the trained deep learning model. Access to these data is provided to the reviewers of the Journal of Water Resources Planning and Management (ASCE) for evaluation purposes. The dataset supports the reproducibility of the proposed deep learning model and facilitates the verification of results reported in the manuscript. File formats: .csv, .xlsx, .pdf, .png, .R and .RmdSoftware used: Python 3.10, TensorFlow, Pandas, NumPy, and AutoCAD Civil 3D for spatial validation.