Cross Entropy Approach to Optimization of the Water Distribution Networks: Optimization of the Hanoi Benchmark Problem

Cross-Entropy (CE) method is a new generic approach for combinatorial optimization and rare event simulation. CE has been applied successfully to optimization problems such as traveling salesman,quadratic assignment problem, the maximal cuts, and partitioning of graphs. The Hanoi problem is one of the problems that are extensively studied by many researchers, using a variety of optimization methods including genetic algorithms, ant colony, simulated annealing and other heuristic methods. In this study, the cross entropy method is applied to the optimization of the Hanoi water distributing benchmark problem.The problem independent nature of CE approach enables ease of use for the optimization of various engineering design problems. The EPANET2 hydraulic solver is linked to the CE Method for optimization of Hanoi bench mark problem. Results show that Hanoi water distributing network benchmark problem could be successfully optimized by CE method with a very limited objective evaluations. It is concluded that CE is not only a robust and easy to use method in dealing with the optimization of the water distribution networks, but also has the capability of rapid convergence to the optimum solutions.

کلیدواژه ها:

Water Distribution Networks ، Optimization ، Visualization ، Self Organizing Feature Maps