A Real-Time Optimal Gate Operation Model for Urban Drainage Systems

Flooding is a phenomenon that endangers human being life and property. There are many structural and non-structural options that can be considered in order to reduce destructive effects of flooding. In this study, we propose a new methodology to enhance the performance of a real-time optimal operation model for flood mitigation in urban drainage systems. An online real-time model is developed as a simulation-optimization approach that leads to optimal operational policies based on the real-time rainfall information. Rainfall-runoff processes and hydraulic routing in the pipes are simulated by the EPA stormwater management model (SWMM) which is linked to the particle swarm optimization (PSO) algorithm, evaluating the system operation performance for assorted sets of operating policies. The initial solution in the real-time model is obtained by a long-term optimal operation model based on historical past flood events. The approach is validated by applying it to a portion of the urban drainage system in Tehran, the capital of Iran, consisting of a prototype network of pipes and detention reservoir equipped with controllable gates. Results show that the proposed strategy in introducing a reasonable initial solution to the real-time model can successfully enhance the performance of the model.