Influência de manobras de válvulas na identificação das rugosidades em tubulações de redes de distribuição de água

Abstract

Currently, based on computational analysis, more accurate assessments have been possible in studies of design of water supply networks. The hydraulic transient is a phenomenon that can affect all forced drainage system. Understanding this phenomenon and the constant search for solutions to avoid its harmful effects cause significant technological advances, principally through the application of computational methods. The present work is to use the Transient Inverse Method (TIM) with a genetic algorithm to calibrate the absolute roughness of pipes of two water distribution networks using four different types of maneuvers through genetic parameters. The study was done in two fictitious networks, but with physical characteristics of real networks. We analyzed the influence of valves maneuvers in the identification of roughness in pipes of water distribution networks. The simulations were made by varying two types of selection: elitism with and without elitism, considering the efficiency indicators: Average Relative Error (EMR), Relative Error (RE) and Objective Function (OF), where the results of the Network proved best for the relative error (RE) with elitism in four maneuvers. In network B, the best indicator of efficiency for the calibration of the four maneuvers was the mean relative error (EMR) with elitism and the indicators Relative Error (RE) and Objective Function (OF) vary according to the analyzed maneuver.