A HYBRID TECHNIQUE FOR OPTIMISATION OF BRANCHED URBAN WATER SYSTEMS

摘要

Evolutionary algorithms (EAs) have demonstrated a unique ability of handling optimization problems. Since they use population based algorithms, they offer a less subjective means of finding many solutions in a single run. Although most of the evolutionary methods can easily converge to the region containing the optimal solution, the solutions found can still be sub-optimal. Evolutionary algorithms have good initial convergence characteristics, but slow down considerably once the optimal region has been identified. This is a major shortcoming of evolutionary algorithms especially in optimization of branched systems, such as water supply, sewer and irrigation systems. In this paper a new methodology is proposed, that combines a local search method with an evolutionary algorithm. The proposed local search strategy (which involves introduction of a set of rules generated based on characteristics of each system and the nature of design constraints) enhances the probability of evolutionary algorithm's convergence to the true optimum. This is a very important improvement in reducing the number of generations and the computational time, which is important in real world problems where cost and time constraints prohibit repeated runs of the algorithm and system simulation. The efficiency of the proposed methodology will be illustrated by application to two branched urban water systems. It will be shown that the method is able to find the optimum solution with less computational effort while satisfying all constraints.