Hasil untuk "River protective works. Regulation. Flood control"

Abdulvahed Khaledi Darvishan, Shahrbanoo Pourbakhshi, Alireza Riyahi Bakhtiari

Understanding sediment sources and budgeting is crucial for effective watershed management and soil conservation. This study employs n-alkanes as biomarkers to trace sediment origins in a small watershed in Northern Iran, comparing the contribution of degraded forests, coniferous afforestation, and natural forests. Soil and vegetation samples were collected from different land use/land covers, while bed sediment samples were obtained from multiple points along the main stream. The distribution patterns of n-alkanes in sediment samples were analyzed in relation to those found in soil and vegetation samples. Additionally, several key indices including the carbon preference index (CPI), aquatic plant proxy (Paq), Hydrocarbon vegetation index (HVI), total organic carbon (TOC), total nitrogen (TN), and soil particle size distribution were measured to further refine source attribution. Using principal component analysis (PCA) and the FingerPro package in R, the contribution of different sediment sources was pinpointed. According to the PCA, the three sediment sources were well separated from each other. The results were striking: degraded forests accounted for the largest contribution at 45.10 %, followed by planted forests at 28.12 % and natural forests at 26.78 %. Considering the area of each land cover, the specific contribution of degraded forest and planted forest to sediment yield were 10.87 % and 7.14 % per hectare, nearly three times and two times that of natural forests (3.78 % per hectare), respectively. Our analysis, validated with a 70 % accuracy rate through the GOF index as well as field evidences, demonstrates that sediment fingerprinting with n-alkanes can effectively reveal erosion patterns and sediment yield rates between different types of forest land use. This insight is crucial for future soil conservation using appropriate afforestation species, ensuring that land management practices are aligned with long-term sustainability goals.

Sadra Shadkani, Yousef Hemmatzadeh, Amirreza Pak et al.

Accurately predicting suspended sediment concentration (SSC) in fluvial systems is essential for environmental monitoring, flood management, and riverine engineering applications. This study introduces a novel hybrid approach for forecasting SSC by leveraging advanced deep learning algorithms. Daily datasets from the U.S. Geological Survey, including discharge (Q) and SSC measurements, were analyzed from 2007 to 2017 at two key locations on the Mississippi River: Chester (CH) and Thebes (TH). The proposed framework integrates feedforward neural networks (FFNN), long short-term memory (LSTM) networks, stochastic gradient descent (SGD), and radial basis function (RBF) models, augmented with a first-order differencing technique. Additionally, hybrid models, including Supervised FFNN-LSTM and Supervised FFNN-SGD, were developed to enhance predictive performance. The dataset was partitioned into training (70%, 2,747 d) and testing (30%, 1,178 d) subsets, with daily temporal resolution. Six input scenarios incorporating lagged parameters were evaluated using performance metrics, including the correlation coefficient (CC), Nash–Sutcliffe efficiency (NSE), scatter index (SI), and Willmott’s index (WI). Sensitivity analysis identified SSCt-1 (i.e., one day before) as the most influential predictor for short-term forecasting. Among the models, the SFFNN-LSTM-6 achieved the highest performance, with CC values of 0.976 for CH and 0.960 for TH, demonstrating the ability to predict SSC effectively even in the absence of current-day discharge data. The proposed hybrid models exhibited exceptional robustness across diverse flow regimes, including extreme environmental conditions, establishing a reliable tool for SSC forecasting in complex fluvial systems.

Jing Guo, Zihan Pan, Xiaomin Yan et al.

Elucidating the historical variation of biogenic elements and source provenance in coastal areas is crucial to better understand environmental evolution and organize ecological management. In this study, the spatiotemporal distribution of biogenic elements (carbon (C), nitrogen (N), phosphorus (P), and silicon (Si)), and carbon and nitrogen isotopes (δ13C and δ15N), respectively in sediment were explored to illustrate elemental changes and determine sources of sedimentary organic matter (SOM) in the coastal regions of northern Beibu Gulf (CNBG). C, N, and P contents exhibited a general decreasing pattern from estuaries to offshore areas, probably due to terrigenous inputs and sediment textures. Since Si was mainly associated with diatoms, lower Si was attributed to the depression of phytoplankton growth by oyster beds in estuaries of Fangcheng Bay and the Maowei Sea. Based on a δ13C and δ15N three-end-member mixing model, it revealed that the average contribution of terrigenous plants and shellfish biodeposition for SOM were 56.00% and 22.39% in the intensive mariculture region (IMR), indicating that terrestrial sources and mariculture biodeposits played a prominent role as SOM sources. Additionally, according to three lead 210 (210Pb) dated sediment cores, it was found that elemental contents and elemental burial rates increased along with the coastal development after 1980s, particularly since 2006 when the Beibu Gulf Economic Zone was established. Multiple Linear Regression results showed that C and N contents and sedimentation rates were responsible for the variation of carbon burial rates during the past decades. More importantly, the significant positive relation between elemental contents and population, and gross domestic product (GDP) growth further proved a close relation between environmental evolution in the CNBG and the socioeconomic development of Guangxi Province. Moreover, the different trends of SOM source contribution for three sediment cores demonstrated the evolution characteristics among different coastal environments. The elevated δ13C and δ15N manifested the increasing contribution of marine phytoplankton (46.83%–69.33%) to SOM in Sanniang Bay recently, where frequent occurrence and decomposition of algal blooms resulted in more SOM. The dominant fraction of terrestrial sources (76.50% ± 13.27%, where ± indicates a standard deviation bound) to SOM in Lianzhou Bay implied the significant impacts of riverine inputs and fishing activities during the last century. This study indicates human activities have led to the continuous increase in nutrients, which has caused ecological risks such as algal blooms in the CNBG, thus, the mitigation of coastal eutrophication needs close attention.

Yunus Ziya Kaya, Fatih Üneş

Abstract Flood modeling is essential to determine and protect vulnerable areas. However, due to complexity of flooding, it is challenging to model floods with a high level of sensitivity. While many factors affect flood models' accuracy, topography is among the most critical. With developing technologies, designing high‐accuracy topographical data is becoming more feasible, especially for small catchments. In this study, the authors focus on macro‐scale modeling using different types of satellite data across the Amik Plain; a large plain with a complex stream network. SRTM, Aster, and Alos Palsar satellite data were used to create digital terrain models (DTMs). The pre‐evaluation of the results showed that even the main streams in the Amik Plain were not visible. So, the geometry of the streams was created and added to the digital elevation models using the HEC‐RAS software RAS Mapper tool. A flood in 2012 was simulated using all three improved DTMs. As a result, it is seen that an enhanced version of the DTM created from SRTM data provides the best performance for use in macro‐scale flood modeling. The usage of the RAS Mapper tool as a GIS tool also performed well in the case of DTM improvements. The DTM improvements on the satellite data for the large plains can give a fairly reasonable output instead of using high‐cost sensitive data.

Joan Estrany, Josep Fortesa, Miquel Tomàs‐Burguera et al.

ABSTRACT Flood events are the most common weather‐related hazard in Europe and Spain, comprising 41% of such events between 2001 and 2020. Mediterranean catchments, with steep slopes and short river courses, are particularly vulnerable to intense convective rainfall, often triggering flash floods. To address this risk, the University of the Balearic Islands developed RiscBal, an innovation ecosystem featuring a high‐resolution Multi‐Hazard Early Warning System. Its core, RiscBal‐Warnings, integrates real‐time data from 56 discharge‐monitoring stations and 32 rainfall/soil moisture stations, forming the RiscBal‐Control network. These stations are positioned in high‐risk and historically flood‐prone areas. This paper focuses on the innovation management behind RiscBal's design and integration into regional governance. Through interdisciplinary collaboration, stakeholder co‐creation, and institutional alignment, RiscBal demonstrates how managing innovation can translate scientific knowledge into actionable, context‐sensitive solutions. The system's performance was tested during the August 15, 2024, flash flood in Es Mercadal, Menorca, providing critical lead time for emergency response. However, issues like telecommunication gaps and early‐stage hydrological modeling prompted improvements, including redundant systems. Riscbal's modular and interoperable design supports polycentric risk governance and continuous feedback between academia, government, and municipalities. Adaptable to other Mediterranean and global flood‐prone regions, it offers a replicable framework for climate resilience. The paper also explores adoption challenges, emphasizing trust, usability and resource constraints.

Hao Hou, Wei Zhu, Cheng Lin et al.

Fluid mud, characterized by a bulk density < 1.2 g/cm3, has various adverse environmental effects; furthermore, it can pose challenges for maritime navigation, as fluid mud is an important factor for determining the depth of navigation channels. Notably, the formation of fluid mud by mixing sand and clay is a complex process. Fluid mud typically occurs on the surface of sediment beds in river estuaries, where the confluence of freshwater and saltwater in such regions promotes particle flocculation. The current study is the first to document the occurrence of fluid mud in Lake Taihu, China. The formation mechanisms were investigated through a comprehensive field and laboratory characterization program. The presence of large flocs was confirmed through this program, prompting further investigation into the role of organic matter. The results indicated that polysaccharides played a crucial role in promoting the aggregation of inorganic particles into flocs. In Lake Taihu, cyanobacterial accumulation zones, which are rich in extracellular polymeric substances (EPS), were primarily found in downwind bays, which also served as the sediment deposition areas. Notably, flocs with a size of ∼80 μm contributed to the formation of highly loose fluid mud in the region, with the density varying from 1.09 to 1.13 g/cm3. Overall, the current study advances the current literature on hydrogeology and sedimentology, particularly with respect to the characterization and formation of fluid mud in lakes.

Mohammad Reza Khazaei

ABSTRACT Overtopping of a dam, which can be due to inadequate spillway design, is often a precursor to dam failure. Generally, dam spillways are designed using historical hydro‐meteorological records. However, the hydrological regime of rivers and floods' intensity may change in the future due to climate change. Although climate change impacts should be considered in spillway design, their consideration remains a challenge. Spillway flow is affected by many factors, each of which can change due to climate change. This paper aims to propose a method for assessing the impacts of climate change on spillway design floods, which considers a wide range of these changes. The method incorporates a robust weather generator, a daily hydrological model, and a reservoir model. The results indicated that the conjunction of the models can acceptably simulate extreme spillway flows. The proposed method was used to assess climate change's impact on the annual maximum spillway flows of a potential large dam in Iran under three Representative Concentration Pathway (RCP) scenarios of the Hadley Centre Global Environment Model version 2‐Earth System (HadGEM2‐ES) model. The findings indicate that while mean monthly spillway flow decreases under all scenarios, extreme spillway flows significantly increase under the RCP 8.5 scenario.

Hakan Varçin, Fatih Üneş, Ercan Gemici et al.

ABSTRACT Dams are built for many purposes, one of the most important of which is to reduce the impact of floods. They release water downstream in a controlled manner through spillway structures. Spillways are critical structures that must be designed with low‐risk criteria to withstand floods that occur once every thousand years. This paper presents a three‐dimensional CFD simulation of the flow over a prototype spillway and validates the results with experimental data. The spillway belongs to the Çatalan Dam, which was constructed in 1985 for hydroelectric power generation. A 1:100 physical model of the Çatalan Dam spillway was built based on Froude similarity, and velocities in the discharge channel were measured at specific cross‐sections and points. The numerical model used the Volume of Fluid (VOF) method and the k‐ε Standard model. The velocity values from the numerical model were compared with experimental velocity values, showing good agreement. Additionally, pressure values were obtained from the numerical model, and cavitation index values were computed accordingly. The results indicated no cavitation risk in the prototype spillway, as the index values remained above 0.2 at all points in the cross‐sections.

Saman Shabani, Yuntong She, Carlos Frederico Lange

The erosion of riverbanks led to significant environmental and infrastructure challenges, leading to soil loss, structural damage, and flooding. Since the 1970s, erosion control structures such as geobags have been widely utilized to protect riverbanks. The geobags provide several advantages, including improved durability and cost-efficiency, compared with traditional rip-raps. To design geobags, it is necessary to understand how the geobags will perform under hydraulic loading. Despite numerous fluvial studies focusing on geobag stability, no study examined how a single bag behaves under current loads, and none of the models simulated the motion of bags. This research investigated geobag movement within the channel bed with overset mesh techniques using computational fluid dynamics (CFD). Additionally, an image processing technique was employed to evaluate the borders of the geobags within the channel. By conducting the simulation, the drag and friction forces acting on the single bag were carefully analyzed, and the effects of the shape, fill ratio and material of the geobags were determined. The failure velocity of geobags following the same pattern was determined based on a correlation with the experiment.

R. S. Wilkho, N. G. Gharaibeh, S. Chang

Abstract Flash floods (FFs) are a leading cause of natural hazard‐related fatalities in the US, posing unique challenges due to their localized impact and rapid onset. Traditional FF susceptibility assessments often fail to account for regional variations. Addressing this, we introduce Dynamic Flash Flood Susceptibility (DFFS), a GIS‐based solution designed for dynamic, region‐specific FF assessment. DFFS operates through four key steps: extracting FF data from the NOAA Storm Events Database for census tracts (CTs) in any region of interest, conducting spatial hotspot analysis to identify areas of high and low FF occurrences, applying causal discovery to identify region‐specific causal factors (from potential factors such as geology, terrain, and meteorology), and using machine learning to calculate susceptibility scores, resulting in a detailed FF susceptibility map. Our case studies in three Texas regions—Dallas‐Fort Worth, Greater Austin, and Greater Houston—revealed distinct causal relationships, with factors like storm duration consistently influential across all regions, while others, such as population density specific to Greater Austin. Furthermore, DFFS demonstrated high accuracy (0.87, 0.86, 0.94) and F1‐scores (0.88, 0.86, 0.96) in computing community susceptibility scores for these regions. We demonstrate DFFS's tangible value in FF risk management and policy‐making, providing a data‐driven and generalizable tool for FF assessment.

Ratnangshu Das, Subhodeep Choudhury, Pushpak Jagtap

This paper presents a spatiotemporal tube (STT)-based control framework for satisfying Signal Temporal Logic (STL) specifications in unknown control-affine systems. We formulate STL constraints as a robust optimization problem (ROP) and recast it as a scenario optimization program (SOP) to construct STTs with formal correctness guarantees. We also propose a closed-form control law that operates independently of the system dynamics, and ensures the system trajectory evolves within the STTs, thereby satisfying the STL specifications. The proposed approach is validated through case studies and comparisons with state-of-the-art methods, demonstrating superior computational efficiency, trajectory quality, and applicability to complex STL tasks.

Jan Olucak, Arthur Castello B. de Oliveira, Torbjørn Cunis

This paper establishes relationships between continuous-time, receding horizon, nonlinear model predictive control (MPC) and control Lyapunov and control barrier functions (CLF/CBF). We show that, if the cost function "behaves well" for points in the terminal set, then the optimal value function and the feasible set, respectively, define a compatible CLF/CBF pair on the MPC's region of attraction. We then proceed to prove that any approximation of the value function and the feasible set also define a CLF/CBF pair, as long as those approximations satisfy the same "well behavedness" condition; and that a feasible state feedback can be computed by solving an infinitesimal version of the MPC problem. This methodology permits the formulation of continuous-time small-sized quadratic programs for feedback and enables approximate solutions of the nonlinear model predictive controller with theoretical safety and convergence guarantee. Finally, we demonstrate the effectiveness of the proposed approach when compared to other constrained control techniques through numerical experiments for nonlinear constrained spacecraft control.

Stefan Ecklebe, Frank Woittennek

This contribution develops an algebraic approach to obtain a controller form for a class of linear hyperbolic MIMO systems, bidirectionally coupled with a linear ODE system at the unactuated boundary. After a short summary of established controller forms for SISO and MIMO ODE as well as SISO hyperbolic PDE systems, it is shown that the direct ap- proach to state a controller form fails already for a very simple MIMO example. Next, a generalised hyperbolic controller form with different variants is proposed and a new flatnesss-based scheme to compute said form is presented. Therein, the system is treated in an algebraic setting where generalised polynomials with real exponents are used to describe the predictions and delays in the system. The proposed algorithm is then applied to the motivating example.

Andreas Paul Zischg

سمیه کریمی اصل, بهزاد حصاری, کامران زینالزاده et al.

مقدمه دشت سلماس، یکی از بحرانی‌‌ترین دشت‌‌های کشور از نظر فرونشست است. بـه‌طورکلی، عوامل مختلفی سبب فرونشست زمین می‌شوند، ولی در بسیاری مناطق برداشت بی‌رویه آب از سفره‌های زیرزمینی باعث فرونشست زمین می‌شود. افزایش روزافزون بهره‌برداری از آب‌های زیرزمینی به‌ویژه در حوضه‌هایی که با نهشته‌های آبرفتی، دریایی کم‌عمق یـا دریاچـه‌ای تحکیم نیافته انباشته شده‌اند، به نشست و یا فروریزش زمین منجر می‌شود. با برداشت بیش از حـد آب زیرزمینی، سطح ایستابی لایه آبدار پایین رفته و فشار هیدرواستاتیک کـاهش مـی‌یابـد کـه درنتیجـه امکان نشست زمین را به‌صورت تدریجی فراهم می‌نماید. فرونشست ماندگار دشـت‌هـا عمدتاً در اثـر همین عامل یعنی برداشت بیش از حد مجاز آب و فشرده‌شدن لایـه‌هـای رسـی و سـیلتی میـان آبخـوان صورت می‌گیرد. در این صورت، حتی اگر سطح ایستابی مجدداً بالاتر آید، امکان بازگشت زمین به تـراز اولیه خود وجود ندارد.   مواد و روش‌‌ها در این پژوهش، حساسیت فرونشست دشت سلماس با بهره‌‌گیری از لایه‌‌های عوامل مؤثر در فرونشست با نرم‌افزار ArcGIS و منطق فازی بررسی شد. در مرحله اول، آمار و اطلاعات برخی عوامل که موجب ایجاد فرونشست می‌‌شوند، شامل افت سطح آب زیرزمینی، دبی بهره‌‌برداری از چاه‏ها، ضریب ذخیره آبخوان، ضریب انتقال‌‌پذیری، بارش، نقشه DEM، بافت خاک و عمق سنگ بستر، جمع‌‌آوری و نقشه رستری هر کدام از این عوامل در سطح آبخوان تهیه شد. در مرحله بعد، فازی‌‌سازی لایه‌‌ها با استفاده از توابع عضویت فازی، با توجه به تأثیر کاهش یا افزایش هر کدام از این عوامل در فرونشست زمین انجام گرفت. سپس، نقشه‌‌ها با عملگرهای فازی (Gamma، OR، AND، SUM، PRODUCT) با هم تلفیق شده و نقشه واحد حساسیت فرونشست آبخوان به‌دست آمد. در نهایت، برای انتخاب بهترین عملگر ترکیب، نتایج را با داده‌‌های مشاهده‌‌ای صحرایی مقایسه شد و با شاخص عملکرد منحنی ROC کنترل و صحت‌‌سنجی صورت گرفت.   نتایج و بحث نتایج نشان داد که عملگر OR با AUC= 0.693 کمترین انطباق را با فرونشست رخ‌داده در منطقه دارد. عملگرهای گاما با سطح زیر منحنی بالای 70 درصد، بیشترین همپوشانی یا انطباق را با فرونشست مشاهده‌‌ای در دشت داشتند و در این پژوهش عملگر گامای 0.9 با  0.805=AUC به‌عنوان بهترین عملگر فازی انتخاب شد. نتایج نشان می‌‌دهد که قسمت شرق آبخوان از نظر فرونشست بحرانی است. از کل مساحت دشت سلماس، 25 درصد معادل 93 کیلومترمربع دارای فرونشست با حساسیت خیلی بالا است.   نتیجه‌‌گیری با توجه به نتایج به‌دست آمده می‌توان بیان کرد که علی‌رغم اینکه مقدار AUC اپراتور فازی sum بالاتر است؛ ولی در نقشه فازی گاما با مقدار 0.9 بیشترین انطباق را با واقعیت زمینی دارد، هرچند مقدار AUC کمتری دارد. ذکر این نکته ضروری است، عملگرهای حداقل‌سازی AND و Product با ایجاد منطقه با حساسیت کم و عملگرهای حداکثرسازی OR و SUM با حداکثرسازی منطقه، نمی‌‌توانند کارایی مطلوبی در تهیه نقشه حساسیت فرونشست داشته باشند و در اینجا فقط برای نمایش‌‌دادن ضعف کارایی اپراتورهای فازی در حداکثر یا حداقل‌سازی حساسیت فرونشست زمین، استفاده شده است.

محمدرضا کوثری, فرود شریفی, علیرضا مجیدی

مقدمه در شرایط تغییر اقلیم و گرمایش جهانی، مدیریت جامع منابع آب و بهره‌وری آن اهمیتی چندین برابر می‌یابد و در این میان اندازه‌گیری منابع آب موجود یکی از پایه‌های اساسی مدیریت منابع آب را تشکیل می‌دهد. اندازه‌گیری هرچه دقیق تر منابع آب موجود، شرایط برنامه‌ریزی بهتر و اساسی را فراهم می‌آورد. آب‌های سطحی خصوصاً منابع آب ناشی از سیلاب‌ها و آورد حوزه‌های آبخیز بزرگ و کوچک، نقش قابل توجهی در چرخه آب ایران بازی می‌کنند. یکی از مشکلات اصلی مدیریت منابع آب در سطح کشور، نبود آمار کافی دبی خروجی از حوزه‌های آبخیز کشور خصوصاً حوضه‌‌های کوچک است. در این میان، استفاده از دستگاه‌های هیدرومتری ثبات و سطح سنجی آب می‌‌تواند به‌طور قابل توجهی مشکلات جمع آوری اطلاعات منابع آب سطحی و زیرزمینی را کاهش دهد. روش‌های متعددی برای اندازه‌‌گیری سطح آب منظور توسعه یافته‌‌اند که می‌‌توان در کل آنها را به دو روش تماسی و غیرتماسی (بسته به قرار گیری حسگر در تماس مستقیم با آب یا خارج از آب) تقسیم‌‌بندی نمود که هر کدام از آنها می‌‌توانند به‌صورت خودکار یا دستی ثبت شوند. انتخاب روش مناسب بستگی به شرایط موجود دارد. برای انتخاب دستگاه سطح‌سنج آب باید مواردی مانند محدوده تغییرات ارتفاع سطح مایع، ویژگی‌های فیزیکی سیال مانند چگالی، تمیز یا کثیف بودن سیال، میزان بخارات یا ذرات جامد موجود در سیال، خورندگی سیال، تمایل سیال به رسوب‌دهی بر روی دیواره ظرف یا وسیله اندازه‌گیری، دما و فشار فرایند، ترکیب شیمیایی سیال، قوانین زیست محیطی و وجود رطوبت، در نظر گرفته شود.   مواد و روش‌‌ها مهم‌ترین مزایای روش‌های غیرتماسی عدم وابستگی زیاد به نوع سیال و عدم درگیر شدن حسگر با سیال است. ازجمله روش‌های غیرتماسی می‌توان به استفاده از عکس‌برداری به‌وسیله دوربین، حسگر آلتراسونیک و همچنین حسگر مادون قرمز و روش لیزری اشاره نمود. در این پژوهش، کارایی یک ماژول مادون قرمز شارپ مدل GP2Y0A02YK0F در اندازه‌گیری تغییرات سطح آب در محیط آزمایشگاه و محیط طبیعی مورد بررسی قرار گرفته است. این ماژول دارای یک حسگر اندازه‌‌گیری فاصله شامل مجموعه‌‌ای یکپارچه از آشکارساز حساس به موقعیت (Position Sensitive Detector, PSD)، دیود مادون قرمز (Infrared Emitting Diode, IRED) و مدار پردازش سیگنال است. محدوده ولتاژ کاری این مجموعه 5/4 تا 5 ولت است. محدوده دمایی عملکرد ماژول بین 10- تا 60+ درجه سانتی‌گراد است. خروجی این ماژول به‌صورت آنالوگ بوده و ولتاژی متناظر با فاصله اندازه‌‌گیری ‌‌شده شامل اعداد در محدوده صفر تا 1023 است. در صورتی که جسم به ماژول نزدیک شود، خروجی به سمت صفر و در حالتی که جسم از آن فاصله بگیرد اعداد خروجی به سمت 1023 میل می‌کند. لذا، نیاز به واسنجی داده‌ها و برقراری ارتباط بین داده‌‌های حسگر و مقادیر واقعی وجود دارد. محدوده اندازه‌‌گیری این سنسور بین 20 تا 150 سانتی‌متر است و از نور مادون قرمز برای تشخیص فاصله استفاده می‌کند. برای بررسی قابلیت‌‌های حسگر در آزمایشگاه و محیط طبیعی، دیتالاگر فوق کم مصرف ساخت داخل کشور که توانایی کار در محیط حوزه‌های آبخیز را داشته باشد مورد استفاده قرار گرفت. البته به علت نیاز به یک سطح غیرقابل عبور جاذب نور، این روش با روش قدیمی استفاده از شناورها بر روی سطح آب ترکیب شد. بعد از اتصال حسگر به دیتالاگر، بدنه مکانیکی دستگاه نیز ساخته شد که شامل یک لوله پلی‌اتیلن است که سنسور در بالای آن قرار گرفته و این مجموعه در داخل یک بدنه فلزی (بدنه مقاوم در برابر سیلاب‌ها و رواناب‌ها) دیگر قرار داده می‌شود. تغییرات سطح آب با استفاده از شناور موجود در لوله و توسط حسگر برداشت می‌شود. این مجموعه ابتدا در آزمایشگاه مورد استفاده و بررسی قرار گرفت. آزمایش اندازه‌‌گیری سطح تغییرات سطح آب در 10 مرحله انجام شد و در هر مرحله تغییرات سطح آب و داده‌‌های حاصل از حسگر برداشت شد. سپس داده‌های برداشت شده حاصل از حسگر به‌عنوان متغیر مستقل و مقادیر واقعی آب به‌عنوان متغیر وابسته در نظر گرفته شد و با برازش چند جمله‌‌ای‌‌ها (از درجه یک تا چهار) معادلات ارتباط بین متغیر مستقل و وابسته تعیین شد. همچنین 30 درصد از داده‌های برداشت شده نیز برای ارزیابی نهایی رابطه‌های ارائه شده در نظر گرفته شد.   نتایج و بحث ارتباط معکوس بین داده‌های حسگر و فاصله واقعی کاملاً مشهود است و با افزایش فاصله مقادیر خروجی حسگر کاهش می‌یابد. مقادیر ضریب همبستگی R در رابطه با برازش‌های یک تا چهارجمله‌‌ای، نزدیک به یک است که نشان‌دهنده همبستگی بالای داده‌‌های حسگر با مقادیر واقعی تغییرات ارتفاع آب است. همچنین، میزان شاخص RMSE نیز بین 16/2 تا  89/1 سانتی‌متر متغیر است.  در کل با افزایش درجه چندجمله‌‌ای، این انطباق بیشتر نیز می‌‌شود و شاخص RMSE نیز کاهش می‌‌یابد. نتایج نشان دهنده آن است که سنسور مذکور در محیط آزمایشگاه با خطای دو سانتی‌متر تغییرات سطح آب را تخمین می‌زند. البته با افزایش دامنه حداقل محدوده اندازه گیری سنسور از 20 به 30 سانتی‌متر، خطای اندازه‌گیری‌ها به 34/1 سانتی‌‌متر کاهش می‌‌یابد. با توجه به قیمت ارزان سنسور، در مواردی که نیازی به دقت بالای اندازه‌‌گیری‌‌ها نیست، این سنسور و روش مناسب است. در صورت نیاز به افزایش دقت، باید حسگرهای دیگر جایگزین شوند. باید در نظر داشت که در هنگام استفاده در محیط‌‌های سیلابی، با مشکلات متعددی ازجمله توقف جنس شناور در داخل لوله اندازه‌‌گیری سطح آب همراه است. لذا، برای اندازه‌‌گیری سطح آب در داخل آبراهه‌های سیلابی توصیه نمی‌‌شود و برای افزایش قابلیت این دستگاه، استفاده از سایر حسگرها مانند حسگرهای فشاری یا التراسونیک، می‌تواند مورد توجه قرار گیرد.    نتیجه‌‌گیری تا به حال روش‌های بسیار زیادی برای اندازه‌گیری تغییرات سطح آب توسعه یافته‌اند، اما آنچه در انتخاب نهایی نوع روش و دستگاه مورد استفاده تاثیرگذار، هدف است. به عبارتی بهتر، متناسب با شرایط محیطی، دقت مورد انتظار و هزینه، نوع دستگاه و حسگر انتخاب می‌‌شود. تا به حال، عملکرد سنسورهای ارزان قیمت GP2Y0A02YK0F در اندازه‌گیری سطح آب گزارش نشده است که در این پژوهش این مهم انجام شد. پژوهش‌های مشابه بیشتر تمرکز بر استفاده از سایر سنسورها خصوصاً سنسورهای فشار یا التراسونیک داشته و البته دیتالاگرها و در کل دستگاه‌های ارائه شده محدودیت تأمین امنیت دارند. با توجه به شرایط فعلی حوزه‌های آبخیز کشور و عدم وجود شبکه‌های گسترده اندازه‌‌گیری سطح آب و دبی، رویکرد کلی ارائه شده در این پژوهش می‌‌تواند، بسیار مؤثر باشد. البته استفاده از شناورها در لوله و بحث نفوذ سیل به لوله یک محدودیت اساسی است و رویکردی بایستی در نظر گرفته شود که استفاده از قطعه شناور را در لوله حذف نماید. چراکه گل آلودگی حاصل از سیل می‌تواند حرکت جسم شناور بر سطح آب را تحت تأثیر قرار دهد و بعد از مدتی از حرکت صحیح و متناسب با سطح آب جسم شناور جلوگیری نماید. در تحقیقات آتی نتایج حاصل از سایر روش‌های اندازه گیری مورد استفاده ارائه خواهد شد.

Changrui Liu, Shengling Shi, Bart De Schutter

Handling model mismatch is a common challenge in model predictive control (MPC). While robust MPC is effective, its conservatism often makes it less desirable. Certainty-equivalence MPC (CE-MPC), which uses a nominal model, offers an appealing alternative due to its design simplicity and low computational costs. This paper investigates CE-MPC for uncertain nonlinear systems with multiplicative parametric uncertainty and input constraints that are inactive at the steady state. The primary contributions are two-fold. First, a novel perturbation analysis of the MPC value function is provided, without assuming the Lipschitz continuity of the stage cost, better tailoring the widely used quadratic cost and having broader applicability in value function approximation, learning-based MPC, and performance-driven MPC design. Second, the stability and performance analysis of CE-MPC are provided, quantifying the suboptimality of CE-MPC compared to the infinite-horizon optimal controller with perfect model knowledge. The results provide insights in how the prediction horizon and model mismatch jointly affect stability and the worst-case performance. Furthermore, the general results are specialized to linear quadratic control, and a competitive ratio bound is derived, serving as the first competitive-ratio bound for MPC of uncertain linear systems with input constraints and multiplicative uncertainty.

Yijiang River Dong, Tiancheng Hu, Nigel Collier

Ensuring that large language models (LLMs) reflect diverse user values and preferences is crucial as their user bases expand globally. It is therefore encouraging to see the growing interest in LLM personalization within the research community. However, current works often rely on the LLM-as-a-Judge approach for evaluation without thoroughly examining its validity. In this paper, we investigate the reliability of LLM-as-a-Personalized-Judge, asking LLMs to judge user preferences based on personas. Our findings suggest that directly applying LLM-as-a-Personalized-Judge is less reliable than previously assumed, showing low and inconsistent agreement with human ground truth. The personas typically used are often overly simplistic, resulting in low predictive power. To address these issues, we introduce verbal uncertainty estimation into the LLM-as-a-Personalized-Judge pipeline, allowing the model to express low confidence on uncertain judgments. This adjustment leads to much higher agreement (above 80%) on high-certainty samples for binary tasks. Through human evaluation, we find that the LLM-as-a-Personalized-Judge achieves comparable performance to third-party humans evaluation and even surpasses human performance on high-certainty samples. Our work indicates that certainty-enhanced LLM-as-a-Personalized-Judge offers a promising direction for developing more reliable and scalable methods for evaluating LLM personalization.

Petar Bevanda, Nicolas Hoischen, Tobias Wittmann et al.

This paper presents a novel operator-theoretic approach for optimal control of nonlinear stochastic systems within reproducing kernel Hilbert spaces. Our learning framework leverages data samples of system dynamics and stage cost functions, with only control penalties and constraints provided. The proposed method directly learns the infinitesimal generator of a controlled stochastic diffusion in an infinite-dimensional hypothesis space. We demonstrate that our approach seamlessly integrates with modern convex operator-theoretic Hamilton-Jacobi-Bellman recursions, enabling a data-driven solution to the optimal control problems. Furthermore, our learning framework includes nonparametric estimators for uncontrolled infinitesimal generators as a special case. Numerical experiments, ranging from synthetic differential equations to simulated robotic systems, showcase the advantages of our approach compared to both modern data-driven and classical nonlinear programming methods for optimal control.

Luke Bhan, Peijia Qin, Miroslav Krstic et al.

Predictor feedback designs are critical for delay-compensating controllers in nonlinear systems. However, these designs are limited in practical applications as predictors cannot be directly implemented, but require numerical approximation schemes, which become computationally prohibitive when system dynamics are expensive to compute. To address this challenge, we recast the predictor design as an operator learning problem, and learn the predictor mapping via a neural operator. We prove the existence of an arbitrarily accurate neural operator approximation of the predictor operator. Under the approximated predictor, we achieve semiglobal practical stability of the closed-loop nonlinear delay system. The estimate is semiglobal in a unique sense - one can enlarge the set of initial states as desired, though this increases the difficulty of training a neural operator, which appears practically in the stability estimate. Furthermore, our analysis holds for any black-box predictor satisfying the universal approximation error bound. We demonstrate the approach by controlling a 5-link robotic manipulator with different neural operator models, achieving significant speedups compared to classic predictor feedback schemes while maintaining closed-loop stability.