Fazal H. Chaudhry

Multiobjective Evolutionary Algorithms Applied to the Rehabilitation of a Water Distribution System: A Comparative Study

Recognising the multiobjective nature of the decision process for rehabilitation of water supply distribution systems, this paper presents a comparative study of two multiobjective evolutionary methods, namely, multiobjective genetic algorithm (MOGA) and strength Pareto evolutionary algorithm (SPEA). The analyses were conducted on a simple hypothetical network for cost minimisation and minimum pressure requirement, treated as a two-objective problem. For the application example studied, SPEA outperforms MOGA in terms of the Pareto fronts produced and processing time required.

Modelos de análise hidráulica de redes de distribuição de água considerando demanda dirigida pela pressão e vazamentos

The hydraulic analysis tools are important in the performance evaluation of water distribution networks. Various methods are available for such analysis. However, the hydraulic models that solve the system of equations describing the flow problem through Newton-Raphson or through its successive linearizations are the most efficient. It is the purpose of this paper to compare Newton-Raphson, Linear Theory, Hybrid and Gradient methods for steady-state hydraulic network analysis, considering leakage and pressure driven demand modeling. The network layout with hydraulic components frequently used in the literature was employed for this analysis. The Gradient method was found to produce best results in terms of number of iterations for the more simple networks, whereas the Hybrid method was better for more complex networks.The hydraulic analysis tools are important in the performance evaluation of water distribution networks. Various methods are available for such analysis. However, the hydraulic models that solve the system of equations describing the flow problem through Newton-Raphson or through its successive linearizations are the most efficient. It is the purpose of this paper to compare Newton-Raphson, Linear Theory, Hybrid and Gradient methods for steady-state hydraulic network analysis, considering leakage and pressure driven demand modeling. The network layout with hydraulic components frequently used in the literature was employed for this analysis. The Gradient method was found to produce best results in terms of number of iterations for the more simple networks, whereas the Hybrid method was better for more complex networks.