Article
Green & Sustainable Science & Technology
Zhenzhen Yue, Wei Huang, Lihua Xiong, Zhuowei Wang, Xuelei Wang, Qian Wang, Qian Shen
Summary: Implementing environmental flow is crucial for river ecosystem restoration and management, but illegal competition for water resources can lead to e-flow shortages. This study developed a monitoring method and guarantee rate evaluation system using remote sensing and hydrological data, showing that e-flow accounted for 24% of average annual runoff. However, errors in the guarantee rate of some stations due to low-resolution remote sensing suggest the need for technological improvements.
JOURNAL OF CLEANER PRODUCTION
(2021)
Review
Chemistry, Multidisciplinary
Yiming Cai, Yao Zhou, Hongwen Zhang, Yuli Xia, Peng Qiao, Junsuo Zhao
Summary: This paper analyzes the calculation principle and error analysis models of aerial camera positioning algorithms and proposes future optimization directions. It provides guidance for researchers to understand the development direction of target geo-location algorithms of aerial remote sensing imagery.
APPLIED SCIENCES-BASEL
(2022)
Article
Engineering, Chemical
Li Wang, Qianhui Tang, Wenhao Li, Xiaoyi Wang, Haiyan Zhang, Jiping Xu, Zhiyao Zhao, Jiabin Yu, Huiyan Zhang, Qian Sun, Yuting Bai
Summary: In this paper, a method of remote sensing image time series pre-processing is proposed, followed by a ACL3DPix2Pix model for pixel-level prediction of remote sensing images. Based on the existing cyanobacterial bloom prediction methods, the spatial and temporal distribution prediction of cyanobacterial blooms is realized by adjusting the existing eutrophication grading criteria. Experimental results demonstrate the effectiveness of this method in predicting cyanobacterial blooms.
DESALINATION AND WATER TREATMENT
(2023)
Article
Engineering, Electrical & Electronic
Jian Pan, Qing Xu, Keqiang Li, Jianqiang Wang
Summary: In this paper, the problem of cloud control of the connected vehicle with time-varying communication-induced sensing delay and control delay is investigated. A discrete linear output feedback system model with bounded input and output delay is proposed for the connected vehicle cloud control system containing communication-induced delay. A predictor-observer structured controller is introduced and redesigned to actively compensate for the delay in both output and input channels. The stability of the control system and a sufficient stability condition are analyzed.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Environmental Sciences
Shaoyan Fan, Ziang Cui, Xuedi Chen, Xinyuan Liu, Fei Xing, Zheng You
Summary: CubeSats have extensive applications in remote sensing, but magnetic field measurement failures can compromise attitude control. To address this issue, we propose a dynamics-sensing, magnetic, fault-tolerant attitude control method that achieves attitude control without a magnetometer and restores remote sensing capabilities.
Article
Green & Sustainable Science & Technology
Xiaojun Zhu, Zhengyuan Ning, Hua Cheng, Pengfei Zhang, Ru Sun, Xiaoyu Yang, Hui Liu
Summary: Coal mining with high groundwater level leads to environmental problems such as surface subsidence and waterlogging, affecting cultivated land and the ecological environment. This study proposes a novel method for calculating the spatial information of subsidence waterlogging using remote sensing technology and surface subsidence prediction, which shows high accuracy and meets engineering precision requirements.
JOURNAL OF CLEANER PRODUCTION
(2022)
Article
Environmental Sciences
Qing Wang, Mengqi Li, Gongquan Li, Jiling Zhang, Shuoyue Yan, Zhuoran Chen, Xiaodong Zhang, Guanzhou Chen
Summary: In this study, a change detection algorithm based on Siamese neural networks is proposed to address the problems of blurred detection boundary, small target miss detection, and more pseudo changes in high-resolution remote sensing image change detection. The algorithm utilizes an improved VGG16 as an encoder to extract image features, and employs an atrous spatial pyramid pooling module to enhance the model's receptive field and obtain multi-scale contextual information. Results from experiments on publicly available CDD and SZTAKI datasets show that the Siam-FAUnet model outperforms baseline models and other state-of-the-art methods, demonstrating its good detection performance.
Article
Environmental Sciences
Zhuowei Wang, Yusheng Lu, Genping Zhao, Chuanliang Sun, Fuhua Zhang, Su He
Summary: This study proposes an advanced biomass estimation approach for sugarcane fields using multi-source remote sensing data, achieving more accurate biomass prediction than traditional linear methods through feature extraction and an integrated regression model. The DAA method extracts a small set of key features, eliminating redundancy in multi-mode data and playing a crucial role in accurate biomass prediction.
Article
Engineering, Civil
Roderick Lammers, Alan Li, Sreeja Nag, Vinay Ravindra
Summary: This research proposes two methods to assimilate satellite-observed precipitation into hydrologic models in real time to update flood forecasts. Using a constellation of small satellites provides frequent flights to capture precipitation data, but requires coordination and planning. Machine learning frameworks show promise in providing accurate flood forecasts with improved computational efficiency.
JOURNAL OF HYDROLOGY
(2021)
Article
Environmental Sciences
Qiang Chen, Qianhao Cheng, Jinfei Wang, Mingyi Du, Lei Zhou, Yang Liu
Summary: With rapid urbanization, the management of urban construction waste is a major concern for urban authorities. The study proposes a multi-feature analysis method for very-high-resolution (VHR) remote sensing images to accurately extract information on urban construction waste. The experimental results show a high identification accuracy for construction waste and demonstrate the effectiveness of the proposed method in resolving spectral confusion between construction waste and surrounding ground objects.
Article
Geochemistry & Geophysics
Jun Chen, Xianqiang He, Wenting Quan, Lingling Ma, Min Jia, Delu Pan
Summary: Residual errors from satellite remote sensing reflectance are influenced by environmental factors, such as image coverage and stray light contamination. Spectral relationships of residual errors should be dynamically reset with the image for accurate data processing.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Meteorology & Atmospheric Sciences
Zheng Duan, Edward Duggan, Cheng Chen, Hongkai Gao, Jianzhi Dong, Junzhi Liu
Summary: Evaluating the accuracy of precipitation products is crucial for many applications, and the multiplicative triple collocation (MTC) method has been shown to be powerful in error quantification when ground truth is unknown. This study applied MTC to evaluate five precipitation products in Germany, revealing that different strategies for replacing zero values can significantly impact MTC-derived error metrics. The MTC-derived correlation coefficient (CC) was found to be more reliable than root-mean-square error (RMSE), making MTC a more suitable method for comparing the relative accuracy of different precipitation products.
JOURNAL OF HYDROMETEOROLOGY
(2021)
Article
Automation & Control Systems
Y. Wardi, C. Seatzu, J. Cortes, M. Egerstedt, S. Shivam, I. Buckley
Summary: This paper presents a control technique for tracking reference signals in continuous-time dynamical systems. The technique uses the fluid-flow version of the Newton-Raphson method, system-output prediction, and a speedup of the control action. It is suitable for linear and nonlinear systems and can track reference points and time-dependent reference signals.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2023)
Article
Engineering, Electrical & Electronic
Ye Gu Kang, David Diaz Reigosa
Summary: This article explores the propagation of current sensor error through the dq-transform and its impact on high-frequency injection (HFI)-based self-sensing control. Statistical models based on variance are developed for 2- and 3-channel-based dq-transform, and their accuracy is verified using error probability density function (PDF) with uniformly distributed random inputs. The results show that the propagated error in the dq-axes is dependent on the rotor position and follows the error variance model. Additionally, the 3-channel based self-sensing control yields lower position estimation error compared to the 2-channel based self-sensing, albeit with the addition of an extra sensor in the machine drive system.
IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS
(2022)
Article
Engineering, Electrical & Electronic
Haoning Lin, Changhao Sun, Yunpeng Liu
Summary: This article proposes a new ensemble method, OBBStacking, which effectively addresses two problems in remote sensing object detection: fusion of oriented bounding boxes and effective use of indicators from deep learning object detectors. The method achieved first place in the 2021 Gaofen Challenge and demonstrated improved performance on multiple datasets.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2023)
Article
Water Resources
Albert Z. Jiang, Edward A. McBean, Andrew D. Binns, Bahram Gharabaghi
Summary: In recent years, flood-related water damages have become the largest home insurance claims in North America. Efforts are being made to reduce flood damages, including reducing sanitary sewer backups that contribute to basement flooding. A guidance methodology on minimum data size requirements has been developed to assess the effectiveness of alternatives, such as reducing rainfall-derived inflows (RDI). The results show that collecting data for approximately 39 storm events provides valuable guidance for field collection programs.
HYDROLOGICAL SCIENCES JOURNAL
(2022)
Article
Computer Science, Interdisciplinary Applications
Iman Ahmadianfar, Mehdi Jamei, Masoud Karbasi, Ahmad Sharafati, Bahram Gharabaghi
Summary: This study combines the prediction powers of Gaussian process regression, random forest, and M5 model tree using a novel ensemble committee-based data intelligent technique to accurately estimate local scour depth around non-uniformly spaced pile groups. The ensemble model significantly outperformed existing empirical models with high correlation coefficient, low root mean square error, and mean absolute percentage of error. Sensitivity analysis showed that pile diameter is the most influential variable in estimating the scour depth.
ENGINEERING WITH COMPUTERS
(2022)
Article
Engineering, Civil
Yar M. Taraky, Yongbo Liu, Bahram Gharabaghi, Edward McBean, Prasad Daggupati, Narayan Kumar Shrestha
Summary: This research examines the impact of headwater reservoirs on climate change and flood frequency in the Kabul River Basin. The study finds that the proposed reservoirs can reduce flooding during the wet season, decrease flood frequency, and increase low flows during the dry season. Additionally, the risks and benefits of reservoirs are discussed in relation to the developmental goals of Afghanistan and Pakistan.
CANADIAN JOURNAL OF CIVIL ENGINEERING
(2022)
Article
Engineering, Civil
K. M. MacKenzie, B. Gharabaghi, A. D. Binns, H. R. Whiteley
Summary: Unmitigated urbanization can lead to the occurrence of urban stream syndrome, which includes increased stream erosion, changes in alluvial materials, and degradation of water quality. This study examines the application of regime-theory equations to identify channels with urban stream syndrome and proposes specific stream power as a reliable early detection metric for this syndrome. The study showcases the use of the GMDH model and a database of channel morphology variables to identify normal and abnormal channels.
JOURNAL OF HYDROLOGY
(2022)
Article
Environmental Sciences
Seyedahmad Kia, Manoj K. Nambiar, Jesse The, Bahram Gharabaghi, Amir A. Aliabadi
Summary: This study predicts methane emissions from a mining facility in Northern Canada using machine learning. By training multiple machine learning algorithms with near-surface observations and weather modeling data from the tailings pond and two open-pit mines, four models were identified as the most accurate in forecasting emissions.
Article
Engineering, Environmental
Ahmed S. Aredah, Omer Faruk Ertugrul, Ahmed A. Sattar, Hossein Bonakdari, Bahram Gharabaghi
Summary: The Extreme Learning Machine (ELM) approach is used to predict stream health and study the influencing factors. The models show good fit and provide better insights on factors influencing stream health, with ELM outperforming other machine learning models.
Article
Construction & Building Technology
Amir Noori, Hossein Bonakdari, Maryam Hassaninia, Khosro Morovati, Iman Khorshidi, Ali Noori, Bahram Gharabaghi
Summary: The growing problem of urban water shortage and its sustainable management methods is a critical research need globally. This study uses GIS and MCDM with triangular fuzzy sets to manage urban water supply priorities in a semi-arid region. A group decision-making approach combining FAHP and FTOPSIS models is proposed based on quantitative and qualitative criteria. A hierarchical model-based GIS and AHP are used to classify effective criteria and determine weights. FTOPSIS is used for priority ranking of scenarios. The study considers environmental, geographical, geological, economic, climatic, and social conditions. A hydrologic model with WEAP software is designed to assess water supply scenarios.
SUSTAINABLE CITIES AND SOCIETY
(2022)
Article
Green & Sustainable Science & Technology
Mohammad Zeynoddin, Hossein Bonakdari, Silvio Jose Gumiere, Alain N. N. Rousseau
Summary: Soil temperature has a significant impact on environmental processes, and this research introduces a technique to address the lag in the FLDAS, which is valuable in watershed-scaled hydrological research.
Article
Agronomy
Guillaume Gregoire, Josee Fortin, Isa Ebtehaj, Hossein Bonakdari
Summary: In this study, a new hybrid machine learning model combining a convolution neural network and a random forest was developed to forecast pesticide use on golf courses in Quebec, Canada. Three groups of variables, including coordinates, characteristics of golf courses, and meteorological variables, were used to estimate pesticide use. The model that considered the latitude and longitude, pesticide type, number of holes, total precipitation, and average temperature from May to November as inputs outperformed other models. The sensitivity analysis indicated that total precipitation was the most critical variable in pesticide use forecasting.
Article
Water Resources
Jean Cardi, Antony Dussel, Clara Letessier, Isa Ebtehaj, Silvio Jose Gumiere, Hossein Bonakdari
Summary: The Ottawa River Watershed is of great importance to Canada, but has experienced increased flooding due to climate change. To accurately predict floods, a combination of numerical modeling and machine learning has been utilized to develop a new ML model for estimating crucial hydrodynamic characteristics of the river.
Article
Environmental Sciences
Victor Oliveira Santos, Paulo Alexandre Costa Rocha, Jesse Van Griensven The, Bahram Gharabaghi
Summary: This study develops a model using graph neural network to predict chloride concentration in Credit River, Canada. The model outperforms other models and shows potential in real-time forecasting of water quality in urban streams.
Proceedings Paper
Instruments & Instrumentation
Amirhossein Zaji, Zheng Liu, Gaozhi Xiao, Pankaj Bhowmik, Jatinder S. Sangha, Yuefeng Ruan
Summary: The goal of this research is to propose a novel object-level data augmentation algorithm that reduces the number of required training samples. The results demonstrate that the algorithm can accurately localize and count wheat spikes, outperforming traditional image-level data augmentation methods.
2022 IEEE INTERNATIONAL INSTRUMENTATION AND MEASUREMENT TECHNOLOGY CONFERENCE (I2MTC 2022)
(2022)
Article
Mathematics, Interdisciplinary Applications
Mansura Jasmine, Abdolmajid Mohammadian, Hossein Bonakdari
Summary: This paper investigates the effective implementation of artificial intelligence on the prediction of evaporation for agricultural area. It presents the adaptive neuro fuzzy inference system (ANFIS) and hybridization of ANFIS with three optimizers. The results suggest that ANFIS and ANFIS-PSO are slightly better than ANFIS-FFA and ANFIS-GA. ANFIS is preferred due to its simplicity and easy operation.
MATHEMATICAL AND COMPUTATIONAL APPLICATIONS
(2022)
Article
Water Resources
Mostafa Elkurdy, Andrew D. Binns, Hossein Bonakdari, Bahram Gharabaghi, Edward McBean
Summary: The study utilized the Generalized Structure Group Method of Data Handling (GS-GMDH) to accurately predict daily and hourly flow data for the Bow River in Alberta, Canada. The model performed well in testing, but issues with horizontal error need to be addressed further.
INTERNATIONAL JOURNAL OF RIVER BASIN MANAGEMENT
(2022)
