Adult American eels (Anguilla rostrata) are vulnerable to hydropower turbine mortality during outmigration from growth habitat in inland waters to the ocean where they spawn. Imaging sonar is a reliable and proven technology for monitoring of fish passage and migration; however, there is no efficient automated method for eel detection. We designed a deep learning model for automated detection of adult American eels from sonar data. The method employs convolution neural network (CNN) to distinguish between 14 images of eels and non-eel objects. Prior to image classification with CNN, background subtraction and wavelet denoising were applied to enhance sonar images. The CNN model was first trained and tested on data obtained from a laboratory experiment, which yielded overall accuracies of >98% for image-based classification. Then, the model was trained and tested on field data that were obtained near the Iroquois Dam located on the St. Lawrence River; the accuracy achieved was commensurate with that of human experts.
An automatic system that utilizes data analytics and machine learning to identify adult American eel in data obtained by imaging sonars is created in this study. Wavelet transform has been applied to de-noise the ARIS sonar data and a convolutional neural network model has been built to classify eels and non-eel objects. Because of the unbalanced amounts of data in laboratory and field experiments, a transfer learning strategy is implemented to fine-tune the convolutional neural network model so that it performs well for both the laboratory and field data. The proposed system can provide important information to develop mitigation strategies for safe passage of out-migrating eels at hydroelectric facilities.
Theoretical studies have shown that cross-correlation functions (CFs) of time series of ambient noise measured at two locations yield approximations to the Green's functions (GFs) that describe propagation between those locations. Specifically, CFs are estimates of weighted GFs. In this paper, it is demonstrated that measured CFs in the 20-70 Hz band can be accurately modeled as weighted GFs using ambient noise data collected in the Florida Straits at ∼100 m depth with horizontal separations of 5 and 10 km. Two weighting functions are employed. These account for (1) the dipole radiation pattern produced by a near-surface source, and (2) coherence loss of surface-reflecting energy in time-averaged CFs resulting from tidal fluctuations. After describing the relationship between CFs and GFs, the inverse problem is considered and is shown to result in an environmental model for which agreement between computed and simulated CFs is good.
Acoustic emissions from current energy converters remain an environmental concern for regulators because of their potential effects on marine life and uncertainties about their effects stemming from a lack of sufficient observational data. Several recent opportunities to characterize tidal turbine sound emissions have begun to fill knowledge gaps and provide a context for future device deployments. In July 2021, a commercial-off-the-shelf hydrophone was deployed in a free-drifting configuration to measure underwater acoustic emissions and characterize a 25 kW-rated tidal turbine at the University of New Hampshire’s Living Bridge Project in Portsmouth, New Hampshire. Sampling methods and analysis were performed in alignment with the recently published IEC 62600-40 Technical Specification for acoustic characterization of marine energy converters. Results from this study indicate acoustic emissions from the turbine were below ambient sound levels and therefore did not have a significant impact on the underwater noise levels of the project site. As a component of Pacific Northwest National Laboratory’s Triton Field Trials (TFiT) described in this Special Issue, this effort provides a valuable use case for the IEC 62600-40 Technical Specification framework and further recommendations for cost-effective technologies and methods for measuring underwater noise at future current energy converter project sites.
Redox flow battery technology has been increasingly recognized as a promising option for large‐scale grid energy storage. Access to high‐fidelity information on the health status of the electrolyte, including the state‐of‐charge (SOC), is vital to maintaining optimal and economical battery operation. In this study, an ultrasonic probing cell that can be used to measure SOC in real time is designed. This unprecedented, new measurement approach overcomes the influence of varying temperatures by measuring the acoustic attenuation coefficient of the redox flow battery electrolyte online and noninvasively. The new approach is used to estimate the SOC of a vanadium redox flow battery in operando from measured acoustic properties. The accuracy of the SOC estimated from the acoustic properties is validated against SOC calculated by the titration method. The results show that the acoustic attenuation coefficient is a robust parameter for SOC monitoring, with a maximum error of 4.8% and extremely low sensitivity to temperature, while sound speed appears to be less accurate in the benchmark‐inference method, with a maximum error of 22.5% and high sensitivity to temperature. The acoustic measurement approach has great potential for inexpensive real‐time SOC monitoring of redox flow battery operations.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.