1,721,135 research outputs found
Fiber-optic sensors for geo-hydrological applications: Basic concepts and applications
No full textIn this short note, basic concepts about fiber-optic sensors are presented, with respect to applications in geology and hydrology. Fiber-Bragg gratings and distributed fiber-optic sensors are then discussed more in detail. © Società Geologica Italiana, Roma 2014
A Review of Distributed Fibre Optic Sensors for Geo-Hydrological Applications
Full text linkDistributed optical fibre sensing, employing either Rayleigh, Raman, or Brillouin scattering, is the only physical-contact sensor technology capable of accurately estimating physical fields with spatial continuity along the fibre. This unique feature and the other features of standard optical fibre sensors (e.g., minimal invasiveness and lightweight, remote powering/interrogating capabilities) have for many years promoted the technology to be a promising candidate for geo-hydrological monitoring. Relentless research efforts are being undertaken to bring the technology to complete maturity through laboratory, physical models, and in-situ tests. The application of distributed optical fibre sensors to geo-hydrological monitoring is here reviewed and discussed, along with basic principles and main acquisition techniques. Among the many existing geo-hydrological processes, the emphasis is placed on those related to soil levees, slopes/landslide, and ground subsidence that constitute a significant percentage of current geohazards
Distributed Sensing in Geotechnical and Hydrological Applications
No full textDistributed Sensing is a disruptive technology in many geotechnical and geo-hydrological applications and, among the different sensing techniques, Optical Frequency Domain Reflectometry can be effectively used at different scales, in laboratory setups, physical models, and in-situ applications
Single-Pump Parametric Amplification in Randomly Birefringent Unidirectionally Spun Fibers
No full textA systematic study of the effects of polarization-mode dispersion (PMD) on broadband and narrowband single-pump fiber parametric amplifiers is realized through numerical solutions of the equations governing the interaction. The nonlinear polarization rotation is shown to be a relevant effect that can increase gain randomness when it mixes with PMD. In unidirectionally spun fibers, the signal-pump alignment, and hence the gain, can be highly enhanced, and nonlinear polarization rotation effects appear for larger PMD values. In spite of the enhanced alignment, for large PMD, the interaction becomes inefficient and the gain decreases to zero
FedZeN: Quadratic Convergence in Zeroth-Order Federated Learning via Incremental Hessian Estimation
No full textFederated learning is a distributed learning framework that allows a set of clients to collaboratively train a model under the orchestration of a central server, without sharing raw data samples. Although in many practical scenarios the derivatives of the objective function are not available, only few works have considered the federated zeroth-order setting, in which functions can only be accessed through a budgeted number of point evaluations. In this work we focus on convex optimization and design the first federated zeroth-order algorithm to estimate the curvature of the global objective, with the purpose of achieving superlinear convergence. We take an incremental Hessian estimator whose error norm converges linearly in expectation, and we adapt it to the federated zeroth-order setting, sampling the random search directions from the Stiefel manifold for improved performance. Both the gradient and Hessian estimators are built at the central server in a communication-efficient and privacy-preserving way by leveraging synchronized pseudo-random number generators. We provide a theoretical analysis of our algorithm, named FedZeN, proving local quadratic convergence with high probability and global linear convergence up to zeroth-order precision. Numerical simulations confirm the superlinear convergence rate and show that our algorithm outperforms the federated zeroth-order methods available in the literature
LQR Temperature Control in smart building via real-time weather forecasting
Full text linkIn this work we consider the problem of climate control within a smart building instrumented with multiple temperature sensors and controllable HVACs. The main contri-bution is to use real-time weather forecasting readily available via internet connection to obtain real-time information about external temperature and solar insulation to reject external disturbances and anticipate temperature changes. The control system is based on MIMO LQR control with a reduced order observer and integral control applied to a model of the building dynamics obtained from construction data. The proposed architecture shows substantial improvements in numerical simu-lations as compared to PID-based standard controllers in terms of improved comfort, while being computationally simpler than more advanced solutions such as MPC or AI-based control
Structural health monitoring of a road tunnel intersecting a large and active landslide
Full text linkDealing with engineering structures that are not easily replaceable requires frequent assessment of the damage state of the construction in order to estimate its durability and reliability. The procedures that allow damage to be detected and identified are broadly defined as Structural Health Monitoring (SHM). In this work, a SHM network has been deployed in a road tunnel that intersects a massive landslide, whose movements are causing the formation of cracks along the tunnel lining. The monitoring system measures in real time the displacements across major cracks and the rotation of the tunnel segments; data are gathered and made easily accessible through a web-based platform. The mechanisms by which the tunnel deforms under the landslide-induced stress have been defined through the analysis of three years of monitoring data. The factors triggering an increase in deformation rates and causing damage to the structure have also been investigated. This evidence will support the design of mitigation works to extend the life-span of the tunnel
Optimal network topology of multi-agent systems subject to computation and communication latency
No full textWe study minimum-variance feedback-control design for a networked control system with retarded dynamics, where inter-agent information exchange is subject to latency. We prove that such a control design can be solved efficiently for circular formations and compute near-optimal control gains in closed form. Also, we show that the centralized control is in general a poor design choice when adding communication links to the network increases the latency, and propose a control-driven optimization of the network topology
Web-Based Platforms for Landslide Risk Mitigation: The State of the Art
Full text linkWeb-based platforms (WBPs) are online spaces where the user can interrogate and analyze data series gathered in quasi-real time from monitoring network/s. These online tools are increasingly used by government agencies, local authorities, contractors, and researchers for visualization, management, control, and analysis of monitoring data. In the risk mitigation framework, WBPs must incorporate specific tools and functions to be integral to non-structural mitigation activities. This is particularly important for the mitigation of landslide-related hazards, that sometimes are challenging to address with structural solutions. The state-of-the-art paper considers the evolution of WBPs for risk mitigation from a pioneering research topic of a decade ago to the current applications that are sometimes comprised within commercial packages. First, we describe what nowadays represents the WBP requirements regarding usability and data visualization for proper data communication. Next, tools for data management and solution regarding interoperability and data analysis are discussed. Lastly, considerations on data filtering in the context of alert and alarm triggering are presented. To be a reasonable alternative to structural mitigation works, non-structural solutions such as monitoring for alarm triggering or early warning must be dependable and stable. We have synthesized the fundamental requirements of monitoring networks devoted to risk mitigation with the expression "5 Rs": robustness, redundancy, reliability, resilience, and responsiveness
- …
