53417 research outputs found
Sort by
Dynamic Wrench-Based Performance Evaluation of an Adjustable Aerial Cable-Towed System
No full textThis paper presents a wrench-based evaluation framework for an adjustable Aerial Cable-Towed System (ACTS), focusing on its capability to generate feasible force and torque combinations under various operational configurations. The ACTS under study consists of three quadrotors, each capable of continuously adjusting the aerial platform’s elevation and azimuth angles. By analyzing the system’s wrench space, its performances are assessed across dynamic trajectory scenario by using capacity margin. The evaluation emphasizes the impact of adjusting the elevation and azimuth angles of the aerial platform to the cable tension while maximizing payload capacity and maneuverability. Results indicate that strategic modulation of the azimuth angle significantly enhances wrench generation capabilities, as shown by increased payload capacities and a higher capacity margin. The dynamic azimuth adjustment effectively minimizes tension loads, validating the potential of wrench-based metrics as a robust tool for optimizing aerial cable-towed operations
Privacy and Energy Aware Deep learning Methods for Energy Theft Detection using IoT Data
Get PDFAbstract Modern smart grid networks, enabled by the Internet of Things (IoT), leverage advanced metering infrastructure to deliver real-time data information from smart meters (SM) to energy companies and consumers. However, IoT data are vulnerable to malicious alterations, potentially leading to energy theft from unbilled consumption. Such thefts can cause billions in annual financial revenue losses. Traditional Machine learning (ML) techniques offer promising solutions for Energy Theft Detection (ETD), yet they struggle with data imbalance and privacy concerns. The data imbalance stems from the dominant representation of honest users and the poor representation of the rare theft cases. Leading ML-based ETD methods employ synthetic data generation to balance the training dataset. However, these methods are trained to maximise average detection instead of ETD, not showing a true reflection of theft detection. To address these issues, we introduce an energy-aware evaluation framework, a convolutional neural network with positive bias (CNN-B) and another with CNN focal loss (CNN-FL), to address the data imbalance by focusing on maximising ETD while minimising revenue loss. As a result, we achieve a revenue loss reduction of 30.4%. Furthermore, this thesis propose a privacy-aware federated learning (FL) based CNN model (FLCNN) that incorporates differential privacy. This model safeguards consumer data during ETD and outperforms traditional centralised methods in terms of bandwidth requirement and reducing false theft attributions. The theft detection of FL-CNN has improved significantly, reducing revenue loss by 41%. Together, these innovations represent a comprehensive approach to enhancing ETD while safeguarding consumer privacy and reducing financial losses in smart grid environments
Genomic and morphometric evidence for Austronesian-mediated pig translocation in the Pacific.
Get PDFSeveral millennia of human-mediated translocation of non-native pig species (genus Sus) to the islands of Wallacea and Oceania have considerably altered local ecosystems. To investigate the timing and trajectory of these introductions, we conducted both genomic analyses of 576 pig nuclear genomes and a geometric morphometric analysis of 708 modern and ancient dental remains. Our analyses demonstrate that free-living and domestic pigs in Wallacea and Oceania have diverse ancestries resulting from the introduction of multiple sequential pig populations followed by gene flow. Despite the variability in their genomic ancestry, these pigs all have a distinct tooth morphology as well as a genetic link to the Chinese domestic pig populations that accompanied the dispersal of Austronesian language speakers ~4000 to 3000 years ago via Taiwan and the Philippines
EXPRESS: Rating traits together or apart: Presentation format affects first impression judgements.
No full textIn perception experiments, researchers often collect multiple perceptual judgments from the same stimuli and participants to answer their research questions. While the way in which perceptual judgements are collected may affect the data, research rarely considers task design when investigating this topic. We therefore investigated how two frequently used ways of collecting multiple ratings for affect perceptual judgements, focusing on first impressions of faces. One participant group provided ratings of seven person characteristics (e.g. femininity, youthfulness, trustworthiness) blocked by characteristic, rating faces for one person characteristic at a time, across seven blocks. Another participant group completed one block in which they rated all seven characteristics simultaneously via a list of rating scales. The listed presentation format reduced task duration by 22%, but affected the perceptual ratings in several ways, pointing to reduced data quality, potentially as a result of satisficing behaviours. Inter-rater agreement was lower for some person characteristics (youthfulness, femininity, and trustworthiness) in the listed format. Variance in ratings was also reduced for youthfulness, femininity, and dominance, with ratings clustering closer to the middle of the scale. Importantly, correlations between different person characteristics became universally positive in the listed format, indicating reduced independence of judgments. These findings highlight that while presenting multiple judgments may offer efficiency, this approach can introduce systematic biases and potentially reduce the reliability of perceptual data. We therefore suggest using a blocked presentation format and consider how these trade-offs would impact experiments looking at multiple perceptual judgements collected from the same participants
Non-Fungible Token (NFT) Relationship with Intellectual Property Rights
No full textNon-fungible tokens (NFTs) are portrayed as a novel system for copyright protection and as exclusive digital assets that represent ownership and originality. However, they do not prima facie have copyright protection, and their relationship with intellectual property (IP) rights is not always straightforward. Despite their trade volume and fluctuating reputation, questions surrounding the legal status of NFTs remain unclear. Current literature on NFTs focuses on either their most common technical aspects or market dynamics, with a minor focus on how different types of NFTs intersect with intellectual property rights. Existing theories often omit the distinction between on-chain and off-chain NFTs, as well as the diverse types of underlying assets. This thesis fills the gap in the literature by analysing the legal nature of NFTs within the framework of the United Kingdom intellectual property law, encompassing a wide range of possibilities and related topics. Among the various intellectual property rights, NFTs most closely resemble copyright. Therefore, this research primarily focuses on analysing copyright jurisprudence. The main argument is that while NFTs cannot be directly equated with copyright, they may qualify for copyright protection under certain conditions. Moreover, this research examines whether the act of minting NFTs can result in infringement and assesses the effectiveness of copyright licensing in NFT transactions. Furthermore, this research aims to clarify the nature of NFT transactions by comparing possibilities and explaining how to classify NFTs as property. This thesis primarily employs doctrinal research, utilising case law, regulatory frameworks, and legal literature and aims to enhance the understanding of the relationship between NFTs and copyright. Addressing gaps in the current literature lays the groundwork for future research and helps define the legal nature of NFTs in relation to copyright and related rights