Genetic algorithm plays a role in municipal water systems

Abstract

Tap water has been taken for granted as a daily resource for many people, but it may not be well known that genetic algorithm optimization has been playing an important role in managing and operating municipal drinking water systems by means of hydraulic and water quality models. Computer models have been built for analyzing water distribution systems since the mid-1960s. The models have become important tools for engineers to simulate various system scenarios. More recently, water quality models have been a critical means for conducting distribution system evaluation studies in order for water companies to comply with USA EPA's latest water quality regulations. However, before hydraulic and water quality models can be used for analysis and operational study of a real system, a model must be able to accurately predict what happens in a real system. The accuracy of modeling results is dependent on a set of model parameters that must be optimized for each water system. Parameter optimization is often referred to as model calibration, that is to minimize the difference between the simulated values and the field observed values. Model calibration is a vitally important process, but a time consuming task.