Maintenance Planning of Water Distribution Networks Based on Multi-Objective Optimization  
Author

Zehui Shi

 

Co-Author(s)

Daqing Gong;  Gang Xue

 

Abstract Frequent failure events in water distribution networks threaten service reliability, increase maintenance costs, and disrupt urban operations. This study proposes a multi-objective repair decision optimization model for sudden failure events. Three sub-models are developed: the GIS indicator extraction model and the EPANET hydraulic simulation model extract and simulate service and traffic-related indicators, which serve as inputs to the impact assessment model for evaluating the priority comprehensive impact (PCI) of failure events; the repair estimation model predicts repair time and material cost for different repair strategies based on historical data. These sub-model outputs are integrated into a genetic algorithm-based multi-objective optimization framework, targeting the minimization of total repair time, total repair cost, and cumulative impact. A case study on a real urban water network validates the approach, showing reductions of 49.7% in repair time, 18.1% in cost, and 57.6% in cumulative impact. The framework provides scientific support for emergency repair planning, enhances water service quality, and improves network reliability.

 

Keywords water distribution networks, multi-objective optimization, simulation, maintenance plan
   
    Article #:  RQD2026-85
 

Proceedings of 31st ISSAT International Conference on Reliability & Quality in Design
August 5-7, 2026