Project Summary


Many engineering design problems can be cast as optimization problems. The Water Industry is not an exception. Water distribution system design is a wide and open problem in hydraulic engineering that involves the addition of new elements in a system; the rehabilitation or replacement of existing elements; decision-making on operation; reliability and protection of the system; among other actions. Designs are necessary in order to carry out new configurations or to enlarge existing ones to meet new conditions.


WaterIng is a software package for water distribution system design and analysis. It offers a multi-objective evolutionary optimization engine based on distributed artificial intelligence to support design and operation decision-making (see below Agent Swarm Optimization). Improved solutions from both hydraulic and economic points of view can be implemented, profiting from the high flexibility of WaterIng platform for defining and evaluating objectives. Different rule-based agents can be incorporated to perform search of solutions more efficiently; depending on the characteristics of the problem, agent population size may vary automatically. WaterIng makes it possible to manage water network data easily by using TableView; additionally, a useful visualization of networks elements over a GIS-based MapView is done. It is possible to import shapefiles, information contained in EPANET files (.INP), or simply copy and paste information from Excel tables to Tableview. Design solutions or existent systems can by analyzed in steady state flow including extended period simulation.

A key point in water distribution system design is to properly assess the objective values of a solution. This is not an easy task, and significant improvements have been obtained in this work by using a recent formulation [1] for evaluating the economy of potential solutions. This formulation makes it possible to embody higher reliability into the water distribution system even under certain failure conditions at reasonable cost. For hydraulic analysis of potential solutions, a pressure driven demand scheme can also be used in order to get closer to reality.


Agent Swarm Optimization: A multi-objective approach.

In general, Optimization problems in the field of urban hydraulics are complex in nature and difficult to solve by conventional optimization techniques. For the last decade, many researchers in the water field have shifted direction, leaving aside traditional optimization techniques based on linear and nonlinear programming and embarking on the implementation of Evolutionary Algorithms: Genetic Algorithms [2-5]; Ant Colony Optimization [6, 7]; Simulated Annealing [8]; Shuffled Complex Evolution [9]; Harmony Search [10]; and Particle Swarm Optimization [11], among others. In water distribution system design, as in many other optimization problems, objectives are frequently in conflict among them. Consequently, it is convenient to find not only one solution but a set of solutions representing the best possible balance regarding the involved objectives. In the last years several proposals have been developed to solve the problem of water distribution system design with a multi-objective approach using evolutionary algorithms [12-14].


In this work a generalization of Particle Swarm Optimization [15] is developed to deal with multi-objective problems and received the denomination of Agent Swarm Optimization (ASO). ASO is oriented to distributed artificial intelligence and profits from the advantages of parallel and distributed computing to ease interaction among several populations of agents with different behaviors. The algorithm offers a common framework to the plurality of existent evolutionary algorithms. Its versatility is the origin of its biggest strength: the introduction of agents with new behavioral rules directly developed to solve a specific problem; these agents can work together with such other evolutionary algorithms as PSO, Genetic Algorithms, Ant Colony Optimization, etcetera. Precisely in water distribution system design, the concept of introducing new agents in the solution process makes it possible that people in charge of projects can also be considered as active agents. Times when experts were just seated in front of computers waiting for results are over. In ASO, human experts are also active agents that can propose solutions in real time and interact with other agents (humans or not) to find better solutions of a problem. Artificial agents can profit from the creativity and ideas of human experts to improve their own solutions; in their turn, human experts can profit from the speed and search capabilities of artificial agents to explore broader solution spaces. In this point ASO makes a difference with respect to classical definitions of multi-agent systems that can be found in the literature.



Development and implementation of a platform for supporting decision making in water distribution system design and operation.


This work is part of a PhD research supported by grant MAEC-AECI 0000202066 awarded to the project manager by the Ministerio de Asuntos Exteriores y Cooperación of Spain.

[1] Martínez, J. B., "Quantifying the economy of water supply looped networks." Journal of Hydraulic Engineering-Asce 133(1), pp 88-97, 2007.


[2] D.A. Savic and G.A. Walters, “Genetic algorithms for least cost design of water distribution networks,” Journal of Water Resources Planning and Management, 123 (2), 67–77, 1997.


[3] Z.Y. Wu and A.R. Simpson, “Competent geneticevolutionary optimization of water distribution systems,” Journal of Computing in Civil Engineering, 15 (2), pp. 89–101, 2001.


[4] A.S. Matías, “Diseño de redes de distribución de agua contemplando la fiabilidad, mediante algoritmos genéticos,” PhD thesis, Universidad Politécnica de Valencia, Valencia, Spain, 2003.


[5] Z. Y. Wu and T. Walski, “Self-adaptive penalty cost for optimal design of water distribution systems,” Journal of Water Resources Planning and Management, 131 (3), 181–192, 2005.


[6] H.R. Maier, A.R. Simpson, A.C. Zecchin, W.K. Foong, K.Y. Phang, H.Y. Seah, and C.L. Tan, “Ant colony optimization for design of water distribution systems,” Journal of Water Resources Planning and Management, 129 (3), pp. 200–209, 2003.


[7] A.C. Zecchin, A.R. Simpson, H.R. Maier and J.B. Nixon, “Parametric study for an ant algorithm applied to water distribution system optimization,” IEEE Trans. Evolutionary Computation, 9 (2), pp. 175–191, 2005.


[8] M.C. Cunha and J. Sousa, “Water distribution network design optimization: simulated annealing approach,” Journal of Water Resources Planning and Management, 125 (4), pp. 214–221, 1999.


[9] S.-Y. Liong and M. Atiquzzaman, “Optimal design of water distribution network using shuffled complex evolution,” Journal of The Institution of Engineers, Singapore 44 (1), pp. 93–107, 2004.


[10] Z. W. Geem, “Optimal cost design of water distribution networks using harmony search,” Engineering Optimization, 38 (3), pp. 259–280, 2006.


[11] Montalvo, I., J. Izquierdo, R. Pérez y P. Iglesias, "A diversity-enriched variant of discrete PSO applied to the design of Water Distribution Networks." Engineering Optimization 40(7), pp. 655-668, 2008.


[12] Vamvakeridou-Lyroudia, L. S., G. A. Walters y D. A. Savic, "Fuzzy Multiobjective Optimization of Water Distribution Networks." Journal of Water Resources Planning and Management 131(6), pp. 467-476, 2005.


[13] Dandy, G. C. y M. O. Engelhardt, "Multi-Objective Trade-Offs between Cost and Reliability in the Replacement of Water Mains." Journal of Water Resources Planning and Management 132(2): 79-88, 2006.


[14] Raad, D., Sinske, A. y J. van Vuuren, "Robust multi-objective optimization for water distribution system design using meta-heuristic."International Transactions in Operational Research Journal Compilation(16), pp. 595-626, 2009.

[15] Kennedy, J. y R. C. Eberhart, Particle swarm optimization. IEEE International Conference on Neural Networks, Perth, Australia, IEEE Service Center, Piscataway, NJ, 1995.