도시 불투수 지역에서 발생하는 초기 강우 유출수는 고농도의 비점오염물질을 단기간에 하천으로 유입시켜 수질 악화를 초래한다. 본 연구에서는 우수받이 및 맨홀에 설치 가능한 다단 폴리프로필렌 섬유 여과장치를 개발하고, 실험실 및 현장자료를 기반으로 오염물질 제거 효율을 평가하였다. 섬유 여재 조합 실험을 통해 53 μm, 20 μm, 10 μm 구성의 3단 여과가 TSS, TN, TP 제거에 가장 효과적인 것으로 나타났다. 현장 초기유출수 분석 결과, 평균 제거 효율은 TSS 79.7%, TN 31.6%, TP 43.1%로 확인되었다. 또한 SWMM을 이용하여 학사마을 도시유역에 적용 시 25%, 50%, 100% 설치비율에 대해 TSS 47.3∼99.2%, TN 21.6∼49.4%, TP 23.6∼67.7%의 저감효과가 모의되었다. 본 연구의 결과는 공간 제약이 큰 도시지역에서 초기세척효과(first flush)를 제어하기 위한 실용적인 비점오염 관리기술로서 활용 가능성을 제시한다.

본 연구에서는 해양플랜트 및 산업용 덕트에서 발생하는 유동 기인 소음을 저감하기 위해 다중공진구조 기반 광대역 소음기를 설계하고, 그 성능을 실험 및 수치해석으로 검증하였다. 제안된 소음기는 공진 특성이 다른 레이어를 주기적으로 배열해 1–3 kHz 구간에 서 국소 공진 차단대역과 Bragg 산란 밴드갭이 동시에 나타나도록 설계되었다. 음향해석 결과, 무유동 조건에서 약 50–60 dB의 삽입손실 이 확보되었으며, 유속 3 m/s와 9 m/s 조건에서도 10–25 dB의 감쇠 성능이 유지되었다. 실험과 수치해석은 전체 주파수 범위에서 정성적 으로 일치했으며, 고속 유동에서의 성능 저하는 공진기 목 주변의 난류 교란과 재생소음(regenerated noise) 영향으로 나타났다. 또한 Bragg 기반 밴드갭은 유동환경에서도 비교적 안정적으로 유지되어 구조적 차단 메커니즘이 성능 저하를 방지하는 데 기여함을 확인하였다. 본 연구는 흡음재 없이도 구조적 설계만으로 광대역 소음 제어가 가능함을 입증했으며, 향후 해양플랜트 및 산업용 덕트 설계 기준 마련에 기여할 수 있다.

본 연구는 광역상수도 요금 변동의 물가파급효과를 정량적으로 분석함으로써 수도 요금 현실화 논의을 위한 기초자료를 제공하고자 한다. 최근 기후위기 심화와 첨단산업의 성장에 따른 용수 수요 증가로 인해 수자원 관리의 중요성이 높아지고 있으며 관련 인프라 투자 확대의 필요성 또한 강조되고 있다. 본 연구는 산업연관분석의 레온티에프 가격모형을 활용하여 광역상수도 요금의 변동(10%)이 국내 생산자물가와 소비자물가에 미치는 영향을 추정하였다. 분석 결과, 생산자물가는 0.0026%, 소비자물가는 0.0033% 수준으로 영향을 주는 것으로 나타나 수도요금 변동이 경제 전반에 미치는 물가영향은 미미한 수준인 것으로 평가된다. 이러한 결과는 광역상수도 요금의 점진적 현실화가 국민 생활과 산업활동에 미치는 영향은 비교적 낮으면서 노후 인프라 개선과 기후변화 대응, 산업용수 수급 안정 등 사회적 필요를 충족시키기 위한 수단으로 기능할 수 있음을 시사한다. 본 연구는 수도 요금 변동의 물가파급효과를 정량적으로 분석함으로써 지속가능한 수도서비스를 위한 재원확보에 대한 기초자료로 활용될 수 있다.

본 연구는 고속도로 다주식 교각 두부보를 대상으로 철근 보강과 철근 대체 GFRP 보강의 균열 손상 거동을 3차원 유한요소 해석으로 비교ㆍ평가하였다. 콘크리트는 ABAQUS의 CDP 모델을 적용하고, 균열 분포는 인장 손상 변수를 핵심 지표로 사용하였다. 선형해석 결과, 두부보 중앙 상단부의 압축 응력 지배 구간과 중앙 하단부 및 접합부 주변의 인장ㆍ전단 영향 구간이 명확히 구분되었 으며, 향후 실험 계측 위치 선정에 활용 가능한 정보를 제공하였다. 비선형 해석 결과, 전반적으로 각 Case의 최초 균열하중 및 최대하 중, 균열 발생 시 변위 및 최대 변위는 큰 차이 없이 유사 범위에 분포하였으며, 균열 면적과 분포 형상 역시 대체로 유사하여 GFRP 보강 두부보의 구조적 안정성이 확인되었다. 특히 일부 Case는 초기 강성과 파괴 저항 측면에서 철근 보강 대비 경쟁력 있는 결과를 보여, 실무적 대체 가능성을 뒷받침하였다.

We present a new fiber assignment algorithm for a robotic fiber positioner system in multi-object spectroscopy. Modern fiber positioner systems typically have overlapping patrol regions, resulting in the number of observable targets being highly dependent on the fiber assignment scheme. To maximize observable targets without fiber collisions, the algorithm proceeds in three steps. First, it assigns the maximum number of targets for a given field of view without considering any collisions between fiber positioners. Then, the fibers in collision are grouped, and the algorithm finds the optimal solution resolving the collision problem within each group. We compare the results from this new algorithm with those from a simple algorithm that assigns targets in descending order of their rank by considering collisions. As a result, we could increase the overall completeness of target assignments by 10% with this new algorithm in comparison with the case using the simple algorithm in a field with 150 fibers. Our new algorithm is designed for the All-sky SPECtroscopic survey of nearby galaxies (A-SPEC) based on the K-SPEC spectrograph system, but can also be applied to similar fiber-based systems with heavily overlapping fiber positioners.

Background: Real-time ergonomic risk assessment in manufacturing environments is challenged by severe class imbalance in high-risk postures and the need for deployment-efficient models. Conventional oversampling techniques may violate biomechanical constraints, limiting their suitability for human motion data. Objectives: This study aimed to compare multiple machine learning models for real-time ergonomic risk assessment while addressing data imbalance using biomechanically appropriate learning strategies and evaluating both predictive performance and deployment efficiency. Design: Comparative study. Methods: A large-scale workplace safety dataset comprising image-based skeletal keypoints was analyzed. To mitigate class imbalance without generating biomechanically implausible samples, cost-sensitive learning and focal loss were employed instead of synthetic oversampling. Subject-wise data splitting was applied to prevent data leakage. Five model families, including Random Forest, convolutional neural networks, and a lightweight graphbased network, were evaluated using accuracy, F1-score, area under the receiver operating characteristic curve (AUC), and high-risk recall. Statistical significance was assessed using bootstrap confidence intervals and McNemar and DeLong tests. Results: The lightweight graph-based model demonstrated competitive classification performance while maintaining reduced computational complexity. Although none of the models achieved the predefined high-risk recall threshold, statistically significant performance differences were observed across model families. Conclusion: The findings suggest that biomechanically informed imbalance handling improves methodological validity in ergonomic risk assessment. While deployment feasibility appears promising, further empirical validation on edge hardware is required.

As the unmanned aerial vehicle industry grows, unexplained multirotor crashes continue to increase, and existing preventive maintenance methods have limitations in managing multirotor safety. Safety must be the top priority in multi-copter operations. To address this, real-time monitoring of the multi-copter's flight status during operation is required, along with anomaly detection and immediate response based on flight log information. However, limitations exist in processing anomaly data for each flight control log, necessitating the development of standardized technology to overcome this challenge. In this paper we propose a standardized process for collecting multi-copter flight control logs in real time, classifying the log information by message sets, and extracting key defect detection indicators contained in each message set. Furthermore, the extracted defect detection indicators were validated using various supervised learning models. In our experimental results, we collected flight logs from a multi-copter equipped with a defective propeller and conducted experiments using three defect detection models. The results show an accuracy rate of 0.99. This is the F1-score for the defect detection rate.

기후변화로 서리의 계절적 발생 시점은 지연되고 있다. 반면에 국내 주요 사과 산지의 서리 발생 빈도는 오히려 증가하고 있어 정밀한 사전 예측의 필요성이 커지고 있다. 본 연구는 노지 과수원 환경을 대상으로, 서리 발생 여부를 예측하는 다중 시간스케일 기반의 인공지능 모델을 제안하였다. 최근 10년간(2014-2025년) 경상북도 안동 기상대의 시간별 관측 97,758건을 사용하였으며, 6·12·24시간의 멀티윈도우 입력으로 단기 급변(복사냉각), 일일 주기성, 장기적인 대기 순환 패턴을 동시에 반영하였다. 모델링은 XGBoost, CNN, XGB-CNN 앙상블로 구성하였으며, 학습-검증-테스트를 70-20-10%로 분할하였다. 성능 평가로 XGB-24h는 ROC-AUC 0.977, PR-AUC 0.921, FPR 0.039로 높은 분별력과 낮은 허위경보를 보였다. CNN-24h는 Recall 0.941로 놓침 최소화에 유리하나 FPR이 상대적으로 높았다. 제안한 앙상블은 두 축을 절충하여 Accuracy 0.932, Recall 0.859, FPR 0.046, MCC 0.809, PR-AUC ≈0.919를 달성했고, Brier 0.056으로 확률 보정도 가장 우수했다. 성능 최적화를 위해 소프트 보팅 앙상블 모델의 가중치(ω)와 서리판정의 임계값(θ)을 대상으로 2차원 grid search를 수행한 결과, 앙상블 성능 조정 시 가중치(ω)보다 임계값(θ)이 핵심 파라미터임을 확인 하였다. 본 연구는 다중 시간스케일과 앙상블에 계절별 동적 임계값(θ) 정책을 적용할 경우 추가적인 성능 개선이 가능함을 시사하며, 지역 일반화의 한계를 고려해 향후 다양한 지역·기후 조건에서의 현장 실증 연구를 통해 재현율(Recall) 중심의 성능향상을 지속적으로 개선하고자 한다.

본 연구는 노인의 소득이 우울에 미치는 영향에서 사회적관계의 매개 효과를 파악하고 이의 경로가 도시와 농촌의 지역별 차이가 있는지 파악 하기 위해 구조방정식 다중집단분석을 수행하였다. 이를 위해 한국복지 패널 19차년도(2024년) 조사에 참여한 65세 이상 노인 6,299명을 연구 대상으로 분석을 수행하였다. 연구 결과, 소득, 사회적관계, 우울의 경로 에는 농촌 노인과 도시 노인 사이에 유의미한 경로차이가 존재하고 있음 이 검증되었다. 특히, 소득에서 사회적관계로 가는 경로와 사회적관계에 서 우울로 가는 경로에서 농촌과 도시의 지역별 유의미한 경로차이가 확 인되었다. 이러한 결과는 도시 노인과 농촌 노인의 지역별 특성을 고려 하여 소득 보장, 사회적관계, 우울 등에 대한 실천적·정책적 개입 방안을 마련해야 함을 강력히 시사하였다.

Enhancing the energy density of electrodes by increasing thickness and mass loading is a technological challenge. Thick electrodes suffer from severe deterioration in electrochemical performance due to insufficient structural integrity and sluggish charge transport, particularly under high current density. Herein, we fabricated thick LiFePO4 (LFP) electrodes with thicknesses ranging from 85.7 to 90.3 μm and an average mass loading of 17.68 mg/cm2 by tailoring the ratio of zero-dimensional (Super P, SP) and one-dimensional (multi-walled carbon nanotube, MWCNT) conductive additives. The electrodes containing MWCNT exhibited crack-free structure and enhanced electrochemical performance with increasing MWCNT ratio because of the superior mechanical properties and electrical conductivity of MWCNT. However, the electrochemical performance of the electrode containing only MWCNT deteriorated due to aggregation of the MWCNT and poor point to point contact with the LFP particles. The multi-dimensional conductive additives improve the dispersion of components within the electrode and the structural stability of the electrode. As a result, the tailored electrode exhibited a lower degree of electrode thickness expansion (1.4 %), lower polarization (60.8 mV at 0.1 C), excellent high-rate capability (132.7 mAh/g at 2 C), superior capacity retention (27.5 % at 3 C), and lower electrical resistivity and interfacial resistance (14.9 Ω cm and 3.8 Ω cm2, respectively) compared to other samples.

Lentic ecosystems, including lakes, reservoirs, and marshes, are vital ecological assets increasingly threatened by anthropogenic pressures, necessitating robust tools for assessing their biological integrity. This study aimed to develop and apply an aquatic plant-based Multi-metric Index (MMI) to evaluate the biological integrity of 90 lentic systems (primarily lakes and reservoirs) across the Republic of Korea, using a standardized dataset from a three-year national monitoring program (2022~2024). We selected eight metrics based on their ecological relevance, sensitivity to disturbance, and scientific robustness. These were organized into three categories: Species richness (30% weight), eutrophication and disturbance (30%), and habitat integrity (40%). Scoring criteria for each metric were established using cumulative distribution functions, and the final MMI scores were used to classify the ecosystems into five integrity classes (A: Excellent to E: Very Poor). The assessment revealed that the majority of the surveyed ecosystems (87.8%) were in a moderate to slightly poor state (Classes B, C, and D), with only 4.4% classified as excellent. Widespread loss of submerged and floating leaved aquatic plants suggests that many domestic lentic systems may be approaching or have already undergone a regime shift to a turbid, phytoplankton-dominated state. This MMI provides a scientifically-defensible tool for managing lentic ecosystems, underscoring the urgent need to restore aquatic plants communities by improving underwater light conditions and rehabilitating littoral habitats.

This study was conducted to develop a fish-based Multi-metric Index (MMI) for assessing the ecological health of lake ecosystems using fish assemblage data collected from the national lake biomonitoring program between 2022 and 2024. A total of 34 fish assessment metrics widely used in the United States, Europe, and Korea were first reviewed for applicability, from which 16 candidate metrics were selected. These candidate metrics were then evaluated in terms of statistical distribution characteristics, correlations with water quality variables, redundancy among metrics, and consistency with existing river-based metrics used in Korea. Based on these evaluations, eight core indicators and four supporting indicators were finalized. For ecological health scoring, boundary values for metric scoring classes were determined using percentiles (10-25-50-75-90%), and metric weights were applied to ensure balanced contribution and discriminative power among classes. The final set of metrics consisted of three indicators in the Diversity/Richness category, two in the Trophic category, two in the Tolerance category, and one in the Individual Health category, collectively reflecting the ecological responses of fish assemblages in lake environments. The developed MMI framework is expected to provide a robust and applicable tool for future ecological assessments and management of lakes in Korea.

Zooplankton are dominant pelagic consumers in lake ecosystems with high population and biomass. Their broad geographical distribution, ease of quantification, and rapid responses to abiotic environmental factors, such as eutrophication, acidification, and climate change, make them highly suitable as indicator organisms for assessing lake ecosystem health. The multi-metric index (MMI) provides an effective framework for capturing the complex responses of biological communities to varying environmental stressors, making it a valuable approach for improving the practical effectiveness of lake ecosystem management based on biological assessments. This study introduces the Lake Zooplankton Assessment Index (LZAI), developed for 90 lakes in South Korea. The LZAI comprises four components: a sensitive species index based on cladocerans, a eutrophication index based on rotifers, a food web index based on copepods, and a habitat index based on species diversity. Applying the LZAI to 90 lakes showed that lake grades followed a normal distribution regardless of sampling season, though A-grade and E-grade lakes exhibited greater seasonal variability. When compared with the clustering results based on zooplankton community composition, the LZAI closely reflected the underlying patterns in community structure. However, in brackish lakes-where population densities are lower and Calanoida copepods dominate relative to freshwater lakes-the M1 and M4 indices were consistently low, while M2 and M3 were high. This suggests that the LZAI requires index adjustments tailored to regional and lake-type factors, including size, depth, and salinity. Incorporating biomass data into the index would further improve the accuracy of assessing community structure and its role in nutrient and energy cycling.

Non-seismic-designed reinforced concrete (RC) pier walls often include lap splices in potential plastic hinge regions. This study develops an analytical model to capture the post-yield load–deformation response of lap-spliced RC pier walls subjected to earthquake loading. The parameters of the model were calibrated using experimental results, and linear regression was conducted to propose predictive equations for these parameters. The accuracy of the model was validated by comparing it to the load–deformation responses of specimens not included in the calibration database. Subsequently, the developed model was applied to probabilistic bridge models supported by RC pier walls. A multi-parameter seismic demand model was constructed, taking into account geometric, material, and structural uncertainties, using Lasso regression. This model achieved R² values of 0.73 for in-plane loading and 0.75 for out-of-plane loading. The improvements in performance metrics and the results of the sensitivity analysis emphasize the need to develop a multi-parameter seismic demand model to ensure more reliable seismic demand predictions.

This study quantitatively analyzed the target strength (TS) characteristics of the dotted gizzard shad (Konosirus punctatus) across various fork lengths (FL) and frequency conditions. In July 2023, TS measurements were conducted on six size groups (FL: 14.4–23.5 cm) under free-swimming conditions in a seawater acoustic tank at the Fisheries Resources Research Center in Tongyeong, Korea. A scientific echosounder (EK80, SIMRAD) was used to collect TS data at three frequencies: 38, 70, and 120 kHz. The results showed that TS values increased with fork length, and the 120 kHz frequency exhibited the widest distribution range and distinct bi- or multi-modal patterns. The TS–FL relationships for each frequency were as follows: TS38 kHz = 20·log(FL) ‒ 68.41, TS70 kHz = 20·log(FL) ‒ 70.76, and TS120 kHz = 20·log(FL) ‒ 70.90. Unlike traditional tethered measurement methods, this study obtained TS data under free-swimming conditions, providing values more representative of real-world acoustic survey environments. The findings are expected to serve as foundational data for improving the accuracy of monitoring the distribution and biomass estimation of K. punctatus using hydroacoustic methods.