Evaluation of a Multi-Objective Genetic Algorithm for Low Impact Development in an Overcrowded City

Abstract

Sustainability and resilience are up-to-date considerations for urban developments in terms of flood mitigation. These considerations usually pose a new challenge to the urban planner because the achievement of a sustainable design through low impact development (LID) practices would be affected by the selection and the distribution of them. This study proposed a means to optimize the distribution of LIDs with the concept of considering the reduction of the flood peak and the hydrologic footprint residence (HFR). The study region is a densely populated place located in New Taipei City. This place has been developing for more than 40 years with completive sewer systems; therefore, the design must consider the space limitations. The flood reduction induced by each LID component under different rainfall return periods was estimated, and then the detention ponds were also conducted to compare the improvements. The results showed that the performance of LIDs dramatically decreased when the return periods were larger than ten years. A multi-objective genetic algorithm (MOGA) was then applied to optimize the spatial distribution of LIDs under different budget scenarios, and to decide the priority of locations for the LID configuration. Finally, the Monte Carlo test was used to test the relationship between the optimal space configuration of LIDs and the impermeability of the study region. A positive correlation was uncovered between the optimal allocation ratio and the impermeable rate of the partition. The study results can provide general guidelines for urban planners to design LIDs in urban areas.