Iranian Jornal of Watershed Management Science&Engineering

fa ارزیابی اثرگذاری روش‌های توسعه کم پیامد بر کنترل رواناب‌های شهری با استفاده از الگوریتم‌های بهینه‌سازی Evaluating the effect of low-impact development methods on urban runoff control using optimization algorithms عمومى General پژوهشي Research <div class="WordSection1" dir="RTL" style="text-align: justify;">در دهه‌های اخیر، توسعه بی‌رویه شهرنشینی، تخریب پوشش گیاهی، تغییر کاربری اراضی موجب افزایش رواناب‌ها و تشدید سیلاب‌های شهری شده است. در این زمینه، مدیریت کمیت و کیفیت رواناب‌ها به منظور کاهش حجم و دبی رواناب‌ها و استفاده مجدد از آنها، اهمیت ویژه‌ای یافته است. در این مقاله، برای اولین بار 16 ترکیب مختلف از هفت روش توسعه کم پیامد ((LID شامل بام سبز، باغچه بارانی، بشکه استحصال باران، ترانشه نفوذ، روکش نفوذپذیر، سامانه نگهداشت زیستی و جوی باغچه با هدف کنترل سیلاب در یک حوزه آبخیز شهری در ناحیه هفت منطقه پنج شهر تهران بکار گرفته شد. مدلسازی هیدرولیکی و هیدرولوژیکی روش‌ها با استفاده از مدل SWMM انجام شد و نتایج با الگوریتم‌های بهینه‌سازی فراابتکاری گرگ خاکستری و ژنتیک رتبه‌بندی نامغلوب در محیط MATLAB تلفیق شد. با توجه به اینکه اکثر مطالعات پیشین بر روی بهینه‌سازی تک‌هدفه روش‌های توسعه کم پیامد تمرکز دارند، این پژوهش یک رویکرد نوآورانه بهینه‌سازی چندهدفه حاصل از ادغام مدل SWMM با الگوریتم‌های NSGAII و GWO را معرفی می نماید که به طور همزمان به توابع هدف مشتمل بر کاهش حجم رواناب، کاهش هزینه روش&lrm;های توسعه کم پیامد و به حداقل‌رساندن کاهش عملکرد روش‌ها در طول دوره بهره‌برداری می‌پردازد. در این پژوهش، روش ترکیبی سامانه نگهداشت زیستی، بام سبز، جوی باغچه و ترانشه نفوذ به عنوان روش ترکیبی بهینه منتخب الگوریتم GWO معرفی شد. ترکیب پیشنهادی الگوریتم GWO با برتری نسبت به الگوریتم دیگر، منجر به کاهش 60-70 درصدی حجم رواناب و کاهش حدود 70 درصدی دبی اوج جریان در خروجی‌های حوضه‌ با هزینه‌ای در حدود 900,000 تا 1,200,000 دلار شد. نتایج این پژوهش قادر است به کارشناسان و مدیران شهری در امور مختلفی مانند طراحی، ارزیابی عملکرد، برآورد هزینه و انتخاب ترکیب‌های بهینه از روش‌های مختلف توسعه کم پیامد یاری ‌رساند.</div> <div style="text-align: justify;"><div aria-label="شکستن صفحه" class="cke_pagebreak" contenteditable="false" data-cke-display-name="pagebreak" data-cke-pagebreak="1" style="page-break-after:always" title="شکستن صفحه"></div></div> Extended Abstract Introduction In recent decades, rapid urban growth, vegetation loss, and land use changes have led to increased runoff and urban flooding. Floods in cities disrupt daily life, challenge urban services, and damage infrastructure. Floods cause significant human and financial losses worldwide, which makes urban runoff and stormwater management a critical challenge. Therefore, effectively managing the quantity and quality of urban runoff has become essential for reducing flooding and facilitating the reuse of water. Low Impact Development methods have shown efficacy in reducing urban runoff, and the main challenge is to propose optimal management measures that decrease runoff volume while exploring the potential use of runoff for other urban purposes. The methods are designed to manage stormwater at its source, minimize runoff, decrease contamination, and provide positive environmental effects. Several studies have shown the benefits of Low Impact Development (LID) methods in managing urban runoff and the other studies have demonstrated their effectiveness when combined with optimization algorithms. This study aims to apply LID methods including green roof, rain garden, rain barrel, infiltration trench, permeable pavement, bioretention system, and vegetative swale and optimize their combinations using metaheuristic algorithms, like the non-dominated sorting genetic algorithm and gray wolf optimizer to control flood in an urban watershed in Tehran. Materials and Methods The study area is located in District 7 and parts of District 1 in Region 5 in the northwest of Tehran, with an area of 9.1 km2. There are two Sub-regions, including 11 sub-watersheds in the study area. The area is a highly urbanized site including residential areas, streets, and parks. This area is especially significant due to the densely populated residential and commercial centers; in fact, it has a great socioeconomic value. It faces flooding and frequent problems in its drainage system because of intense rainfall events, inadequate channel capacities, and urban waste repletion. For the first time, this research utilized sixteen various combinations of 7 LID methods, including green roof, rain garden, rain barrel, infiltration trench, permeable pavement, bioretention system, and vegetative swale to control floods in the urban watershed. Hydraulic and hydrological modeling were conducted using a six-hour design rainfall, and the parameters for LID methods settlement were established in the SWMM model for the 25-year return period. The results were integrated and evaluated using NSGA-II and GWO in MATLAB. The answer on the Pareto fronts generated by NSGA-II and GWO algorithm were compared and the optimized answers were proposed by the GWO algorithm. The innovative approach of this research simultaneously addresses runoff reduction, cost efficiency, and service performance sustainability. Introducing a desirable optimization algorithm, it proposed a suitable optimized combination for urban flood control in the study area. Results and Discussion Calibration and validation of the SWMM model demonstrated its accurate performance in simulating floods in this area. The results of simulating total runoff and peak discharge without implementing LID methods showed that the drainage network is not suitable for rainfall with a 25-year return period, and it is unable to transfer the runoff properly. Modeling LID methods for 25-year, 6-hour rainfall, separately and in various combinations, indicated that the greatest effectiveness of LIDs on reducing runoff volume and discharge is associated with combinations of three or four methods. Running the coupled model of SWMM with NSGA-II and GWO algorithms for various combinations of LID methods and comparing the results, the combination of bioretention system, green roof, vegetative swale, and infiltration trench was chosen because it could effectively reduce runoff volume, be cost-effective, and demonstrate minimal performance reduction compared to other combinations. Comparing the answers provided by the optimization algorithms, the Pareto fronts generated by the GWO algorithm suggested suitable solutions for various values of all three objective functions, due to their better density and distribution. Therefore, the optimized combination proposed by the GWO algorithm, demonstrated the best performance by surpassing another method, and could reduce runoff volume by 60 to 70 percent and peak discharge by 70 percent in the study area. The cost of implementing the combination is approximately $900,000 to $1,200,000. Conclusion The growth of urban areas and the transformation of natural landscapes into cities have resulted in more impervious surfaces, increased runoff, and a higher risk of flooding. Low Impact Development (LID) methods have been suggested as effective solutions to address these challenges. This study conducted multi-objective optimization of LID methods in an urban watershed using the SWMM model and the NSGA-II and GWO algorithms. Unlike previous studies that mainly emphasized single-objective optimization, this research simultaneously focused on minimizing runoff volume, minimizing cost, and the sustainability of methods’ performance. The findings showed that the combination of bioretention systems, green roofs, vegetative swales, and infiltration trenches is very effective in controlling runoff and flooding in the study area. The GWO algorithm is more efficient compared to NSGA-II, achieving a significant reduction of 60 to 70 percent in runoff volume at a reasonable cost. The proposed solutions of this research can play a crucial role in reducing water loss in arid and semi-arid regions, thereby facilitating the optimal utilization of water resources. Ultimately, the simulation-optimization approach serves as a reliable and effective tool for urban planning experts and flood control specialists to improve the quality and quantity of urban water by selecting the best combination of LID methods. الگوریتم بهینه‌سازی گرگ خاکستری, الگوریتم ژنتیک رتبه‌بندی نامغلوب, روش‌های توسعه کم پیامد, کنترل سیل, مدل مدیریت آب باران Flood Management, Gray Wolf Optimizer, Low impact development methods, Non-dominated sorting genetic algorithm II, Storm Water Management Model 52 71 http://jwmsei.ir/browse.php?a_code=A-10-121-4&slc_lang=fa&sid=1 Sedigheh Modarresi Tabatabaei صدیقه مدرسی طباطبایی hedyeh_modarres@yahoo.com 100319475328460019165 0009-0003-4224-5866 No Tarbiat Modares University دانشگاه تربیت مدرس Hamidreza Moradi حمید رضا مرادی hrmoradi@modares.ac.ir 100319475328460019166 100319475328460019166 Yes Tarbiat Modares University دانشگاه تربیت مدرس Mehdi Vafakhah مهدی وفاخواه vafakhah@modares.ac.ir 100319475328460019167 100319475328460019167 No Tarbiat Modares University دانشگاه تربیت مدرس Javad Vahidi جواد وحیدی jvahidi@iust.ac.ir 100319475328460019168 100319475328460019168 No Iran University of Science and Technology دانشگاه علم و صنعت