이병철, 강정은. (2026). 연안침식 후퇴구역 설정을 위한 개미 군집 최적화 알고리즘의 활용. 2026 대한국토도시계획 춘계산학학술대회. (2026. 04. 23. - 2026. 04. 24.)  

[Abstract] 

  This study aims to compare and evaluate the performance of strategies for establishing uniform setback zones based on distance and selective setback zones at the parcel level. This research is conducted against the backdrop of intensifying coastal erosion on sandy beaches along the East Coast due to sea-level rise and increased extreme waves caused by climate change, and the growing importance of setback zones as a long-term adaptation strategy to address these challenges. Using Baekseok Beach in Yeongdeok County, North Gyeongsang Province as a case study, the Beach Hardness Assessment model was applied to predict shoreline retreat and coastal erosion risk zones through 2100 for each SSP scenario based on IPCC AR6. Within these risk zones, lots were used as the unit of analysis to construct economic, social, and environmental indicators, including land and building acquisition costs, resident and worker populations, and green space area. Subsequently, under the same budget constraints (at the level of existing coastal erosion prevention and beach nourishment project costs), two distance-based retreat area scenarios were established based on a fixed distance from the coastline, varying only by the inclusion of buildings. Additionally, two optimization-based retreat area scenarios were developed using Ant Colony Optimization (ACO) to consider economic, social, and environmental exposures on a multi-criteria basis for each plot, and their performance was compared. The analysis revealed that while the passive distance-based scenarios had lower costs, they included almost no residents, workers, or buildings within the setback areas, resulting in negligible practical risk reduction effects. In contrast, the active distance-based scenarios concentrated the budget on coastal commercial and accommodation zones, primarily protecting workers and buildings; however, under the same budget constraints, they were less efficient than the ACO-based scenarios in terms of protecting residents and conserving green spaces. In contrast, ACO-based selective retreat areas incorporated more residents and green spaces while maintaining the budget cap. In particular, the scenario that weighted social vulnerability showed the highest number of protected residents per unit cost, and was evaluated as a strategy aligned with the policy goal of prioritizing the protection of vulnerable groups. By presenting a data-driven setback area design framework that combines coastal erosion prediction with a meta-heuristic-based land selection model, this study is expected to contribute to lot-level selective setback area design and budget allocation decision-making in future coastal, national, and urban planning, thereby complementing distance-based uniform regulations. 

사사 : 연안침식