Document Type : Research Paper

Authors

• Saeed Amiry
• Nazila Kardan
• Mehdi Dini

Department of Civil Engineering, Azarbaijan Shahid Madani University, Tabriz, Iran.

Abstract

Urban stormwater drainage networks in rapidly growing cities often suffer from undersized infrastructure, leading to frequent flooding and high rehabilitation costs. While evolutionary algorithms have been widely applied to optimize pipe diameters and layouts, existing approaches typically employ single-objective or global search methods, neglecting local topographic constraints and real-scale validation with high-resolution data. This study proposes a novel two-stage hybrid framework integrating SWMM, GIS, Cellular Automata (CA), and Genetic Algorithm (GA) for optimal redesign of urban surface water collection networks. In Stage 1, CA performs local refinement of node burial depths using neighborhood rules to minimize excavation costs while ensuring favorable hydraulic slopes. The resulting fixed vertical layout is then passed to Stage 2, where GA globally optimizes discrete pipe diameters and routing to minimize total construction costs under penalized hydraulic constraints, with dynamic evaluation. Applied to a real-world case study in District 7 of Tabriz, Iran (112 nodes, 168 pipes, 12.3 km²), the framework incorporates high-resolution DTM (±5 cm accuracy) and field-surveyed infrastructure data. Results demonstrate elimination of flooding under a 25-year design storm and a 32% reduction in total cost compared to conventional manual design, yielding practical recommendations including channel widening, dredging depths, and prioritized interventions. By decoupling local and global optimization and leveraging detailed topographic integration, the proposed methodology advances current stormwater network optimization practices, offering improved cost-efficiency, hydraulic performance, and applicability to large-scale urban systems.

Keywords

• Urban runoff management
• SWMM
• Cellular Automata
• Genetic Algorithm