Fuzzy Control of Pumps in a Combined Sewer Pumping Station Using Genetic Argorithms

Abstract

To achieve advanced control in sewer systems, operators should be provided with computer-aided control systems. In the present study, genetic algorithms were applied to the learning of fuzzy control rules used in a combined sewer pumping station. Pumping rates were determined by fuzzy logic controllers using input variables such as rainfall intensity, sewer water level, and river water level. Genetic algorithms involve the procedures of selection, crossover, and mutation. After evaluating the effects of parameters, fuzzy control rules were efficiently improved to fit pump operation conducted by human operators.