This study proposes a novel simulation-optimization approach using WaterGEMS and evolutionary hybrid algorithms (ANFIS-PSO and ANFIS-GA) alongside general ANFIS to predict water distribution system performance. Two water distribution networks (WDN), Go-Yang and Gadhra WDN are considered. Velocity prediction depends on pipe diameter and length, while pressure prediction depends on node elevation and demand. Statistical analysis of the three hybrid models revealed that ANFIS-PSO outperformed ANFIS-GA and general ANFIS in velocity and pressure predictions for both WDNs. In the case of Go-Yang WDN, ANFIS-PSO achieved a coefficient of determination (R2) of 0.9884, a root mean square error (RMSE) of 0.0353, and a mean absolute relative error (MARE) of 0.1731 for velocity and ANFIS-PSO achieves (R2, RMSE and MARE) as (0.9812, 0.7256 and 0.03) respectively for pressure. In case of Gadhra WDN, ANFIS-PSO shows the statistical performance (R2, RMSE, and MARE) as (0.9143, 0.0441 and 0.6684) respectively for velocity and ANFIS-PSO achieved (R2, RMSE and MARE) as (0.8929, 0.3875, and 0.02) respectively for pressure. The computational time required for ANFIS is less than 1 min whereas in the case of ANFIS-evolutionary techniques (ANFIS-GA and ANFIS-PSO) it is about nine minutes.
Simulation-Optimization of Water Distribution Networks Using ANFIS-Evolutionary Techniques
KSCE J Civ Eng
KSCE Journal of Civil Engineering ; 28 , 1 ; 484-494
2024-01-01
11 pages
Article (Journal)
Electronic Resource
English
Automated Aerodynamic Vehicle Shape Optimization Using Neural Networks and Evolutionary Optimization
SAE Technical Papers | 2015
|Automated Aerodynamic Vehicle Shape Optimization Using Neural Networks and Evolutionary Optimization
British Library Conference Proceedings | 2015
|Automated aerodynamic vehicle shape optimization using neural networks and evolutionary optimization
Automotive engineering | 2015
|Comparison of modified Karnik-Mendel algorithm-based interval type-2 ANFIS and type-1 ANFIS
Emerald Group Publishing | 2021
|