MRC Laboratory of Molecular Biology
CUED - Cambridge University Engineering DepartmentNot a member yet
45551 research outputs found
Sort by
Experimental investigation on scour topography around high-rise structure foundations
No full textThe current study aims to investigate the characteristics of scour topography around High-Rise Structure Foundations (HRSFs) via physical modeling tests. Clear-water scour tests with a uniform non-cohesive bed are modeled under the action of unidirectional steady flows. Time variations of the erosion and deposition topography are measured. The results show that deposition downstream of the first dune behind the HRSF is not located on the centerline of the wake. The deposition pattern indicates that a long steady wake region exists behind the permeable foundation. The scour depth around an HRSF is much less than that around a monopile because of the structural permeability, which gives rise to the bleed flow and a weakened downflow and horseshoe vortex. Additionally, the asymmetry of the HRSF affects the scour rate but not the final equilibrium scour depth. The average scour slope decreases along the direction of the flow. On the contrary, the scour radial distance increases along the direction of the flow, with the average value changing from 1.36De to 2.35De (where De is the equivalent diameter of the foundation). Furthermore, the scour hole around the HRSF is serrated rather than smooth owing to the presence of multiple piles. Empirical formulae are suggested for estimating the evolution of scour depth and volume. These laboratory experiments provide reference information for relevant numerical modeling studies and can be applied to guide engineering designs in an ocean area
Self-supervised transfer learning of physiological representations from free-living wearable data
No full textWearable devices such as smartwatches are becoming increasingly popular tools for objectively monitoring physical activity in free-living conditions. To date, research has primarily focused on the purely supervised task of human activity recognition, demonstrating limited success in inferring high-level health outcomes from low-level signals. Here, we present a novel self-supervised representation learning method using activity and heart rate (HR) signals without semantic labels. With a deep neural network, we set HR responses as the supervisory signal for the activity data, leveraging their underlying physiological relationship. In addition, we propose a custom quantile loss function that accounts for the long-tailed HR distribution present in the general population. We evaluate our model in the largest free-living combined-sensing dataset (comprising >280k hours of wrist accelerometer & wearable ECG data). Our contributions are two-fold: I) the pre-training task creates a model that can accurately forecast HR based only on cheap activity sensors, and ii) we leverage the information captured through this task by proposing a simple method to aggregate the learnt latent representations (embeddings) from the window-level to user-level. Notably, we show that the embeddings can generalize in various downstream tasks through transfer learning with linear classifiers, capturing physiologically meaningful, personalized information. For instance, they can be used to predict variables associated with individuals' health, fitness and demographic characteristics (AUC >70), outperforming unsupervised autoencoders and common bio-markers. Overall, we propose the first multimodal self-supervised method for behavioral and physiological data with implications for large-scale health and lifestyle monitoring. Code: Https://github.com/sdimi/Step2heart
A data-driven dynamic repositioning model in bicycle-sharing systems
No full textThe new generation of bicycle-sharing is an O2O (online-to-offline) platform service that enables the users to access the bicycle with a smartphone App. This paper proposes a dynamic repositioning model with predicted demand, where the repositioning time interval is fixed. A data-driven Neural Network (NN) approach is introduced to forecast the bicycle-sharing demand. The repositioning objective function at each time interval is defined to simultaneously minimize the operator cost and penalty cost. In addition to the normal constraints in static repositioning problem, flow conservation, inventory-balance and travel time constraints are taken into account. Due to the non-deterministic polynomial-time hard (NP-hard) nature of this model, a hybrid metaheuristic approach of Adaptive Genetic Algorithm (AGA) and Granular Tabu Search (GTS) algorithm is applied to calculate the solution. Based on predicted demand, the initial repositioning plan is made by AGA statically at the beginning of study horizon, which ensures the global optimization of the first solution. As time goes on, repositioning plan is checked and updated according to the real-usage patterns using GTS algorithm, which has the advantage of high-performance local-search within a short computing time. Numerical analysis is conducted using the real cases. The simulation results reveal that the proposed methodology can effectively model the dynamic repositioning problem in response to real-time bicycle-sharing usage. The proposed methodology can be a value-added tool in enhancing the feasibility and sustainability of bicycle-sharing program
Chapter 13: Polymers/PEDOT Derivatives for Bioelectronics
No full textThe advancement of bioelectronics depends greatly on new material development and engineering solutions. Redox polymers are promising candidates to contribute to this advancement of biointerfacing devices. For such devices to be clinically useful, they must fulfill an assortment of requirements, including biocompatibility, stability, mechanical compliancy and the ability to effectively monitor or influence biological systems. The use of redox polymers in bioelectronic research has demonstrated a great deal of potential in satisfying these constraints. In this chapter, we consider the advantageous aspects of polymer electronics for biomedical applications including electrophysiological recording, neuromodulation, biosensor technologies and drug delivery. Particular emphasis is given to PEDOT-based systems as these have demonstrated the highest degree of bioelectronic device success to date, however, other polymers are also discussed when pertinent
Interwoven scaffolded porous titanium oxide nanocubes/carbon nanotubes framework for high-performance sodium-ion battery
No full textSupercapacitor-like Na-ion batteries have attracted much attention due to the high energy density of batteries and power density of capacitors. Titanium dioxide (TiO2), is a promising anode material. Its performance is however seriously hindered by its low electrical conductivity and the sluggish diffusion of sodium ions (Na+) in the TiO2 matrix. Herein, this work combines porous TiO2 nanocubes with carbon nanotubes (CNTs) to enhance the electrical conductivity and accelerate Na+ diffusivity for Na-ion batteries (NIBs). In this composite, an interwoven scaffolded TiO2/CNTs framework is formed to provide abundant channels and shorter diffusion pathways for electrons and ions. The in-situ X-ray diffraction and cyclic voltammetry confirm the low strain and superior transport kinetics in Na+ intercalation/extraction processes. In addition, the chemically bonded TiO2/CNTs hybrid provides a more feasible channel for Na+ insertion/extraction with a much lower energy barrier. Consequently, the TiO2/CNTs composite exhibits excellent electrochemical performance with a capacity of 223.4 mAh g−1 at 1 C and a capacity of 142.8 mAh g−1 at 10 C (3.35 A g−1). The work here reveals that the combination of active materials with CNTs can largely improve the utilization efficiency and enhance their sodium storage
Optimal quantization for amplitude and phase in computer-generated holography
No full textOwing to the characteristics of existing spatial light modulators (SLMs), the computergenerated hologram (CGH) with continuous complex-amplitude is conventionally converted to a quantized amplitude-only or phase-only CGH in practical applications. The quantization of CGH significantly affects the holographic reconstruction quality. In thiswork, we evaluated the influence of the quantization for both amplitude and phase on the quality of holographic reconstructions by traversing method. Furthermore, we considered several critical CGH parameters, including resolution, zero-padding size, reconstruction distance, wavelength, random phase, pixel pitch, bit depth, phase modulation deviation, and filling factor. Based on evaluations, the optimal quantization for both available and future SLM devices is suggested
Large deformation analysis of spontaneous twist and contraction in nematic elastomer fibers with helical director
No full textA cylindrical rubber fiber subject to a twist will also elongate: a manifestation of Poynting's effect in large strain elasticity. Here, we construct an analogous treatment for an active rubber fiber actuated via an axisymmetric pattern of spontaneous distortion. We start by constructing an exact large-deformation solution to the equations of elasticity for such fiber subject to imposed twist and stretch, which reveals spontaneous warping and twisting of the fiber cross section absent in passive rubbers. We then compute the corresponding non-linear elastic energy, which encompasses the Poynting effect but is minimized by a finite spontaneous twist and stretch. In the second half of the paper, we apply these results to understand the twist-contraction actuation of nematic elastomer fibers fabricated with director fields that encode helical patterns of contraction on heating. We first consider patterns making a constant angle with respect to the local cylindrical coordinate system (conical spiral director curves) and verify the predicted spontaneous twist, contraction, and cross-section deformation via finite elements. Second, we consider realistic director distributions for the experimentally reported fibers fabricated by cross-linking while simultaneously applying stretch and twist. Counterintuitively, we find that the maximum actuation twist is produced by applying a finite optimal twist during fabrication. Finally, we illustrate that spontaneously twisting fibers will coil into spring-like shapes on actuation if the ends are prevented from twisting relative to each other. Such a twist-torsion coupling would allow us to make a tendril-like "soft-spring"actuator with low force and high linear stroke compared to the intrinsic contraction of the elastomer itself
Relaxing Platform Dependencies in Agent-Based Control Systems
No full textAgent-based systems have been widely used to develop industrial control systems when they are required to address issues such as flexibility, scalability and portability. The most common approach to develop such control systems is with agents embedded in a platform that provides software libraries and runtime services that ease the development process. These platforms also bring challenges to the agent-based control system engineering. For example, they might introduce default design features, such as a global directory of agents. Furthermore, agents are generally locked in a platform and depend on the platform's available support for deployment across computing infrastructures. This article addresses these challenges through an approach for building agent-based control systems, that relaxes the dependencies in multiagent system (MAS) platforms, through the use of container-based virtualisation. The proposed approach is elaborated via a reference architecture that enables the implementation of agents as self-contained applications that can be deployed, on-demand, in independent environments but still are able to communicate and coordinate with other agents of the control system. We built a prototype using this approach and evaluated it in the context of a case study for the supervisory control of digital network infrastructures. This case study enabled us to demonstrate feasibility of the approach and to show the flexibility, of the resulting control system, to adopt several topologies as well as to operate at different scales, over emulated networks. We also concluded that designing agents as individual deployment units is also cost-effective especially in control scenarios with low number of stable agents
Coupled VO<inf>2</inf> Oscillators Circuit as Analog First Layer Filter in Convolutional Neural Networks
No full textIn this work we present an in-memory computing platform based on coupled VO2 oscillators fabricated in a crossbar configuration on silicon. Compared to existing platforms, the crossbar configuration promises significant improvements in terms of area density and oscillation frequency. Further, the crossbar devices exhibit low variability and extended reliability, hence, enabling experiments on 4-coupled oscillator. We demonstrate the neuromorphic computing capabilities using the phase relation of the oscillators. As an application, we propose to replace digital filtering operation in a convolutional neural network with oscillating circuits. The concept is tested with a VGG13 architecture on the MNIST dataset, achieving performances of 95% in the recognition task
The Effect of Cross-Flow Vortex Trap Devices on the Aerodynamic Drag of Road Haulage Vehicles
No full textThe effect of Cross-Flow Vortex Trap Devices (CVTDs) on the local flow field and vehicle drag at a range of yaw angles has been investigated in wind tunnel experiments. The CVTD is a flow control device proposed by Bauer and Wood that aims to reduce the sensitivity of articulated road haulage vehicles to crosswinds by managing the tractor-trailer gap cross-flow. A 1/10th scale model is used in a low-speed wind tunnel at a Reynolds number of 900,000. The aerodynamic drag force is measured using a load cell connected to a rotating, raised ground plane. This research also uses tuft flow visualization to examine the local flow fields and pressure taps to determine trailer pressure distributions. It is found that a configuration of four 45% length CVTDs reduces the wind-averaged drag coefficient by 12%. The drag mechanisms responsible for the reduced drag include a lower average pressure on the trailer front face, removal of the separation on the leeward side of the trailer due to a reduction in the gap cross-flow, and an increase in pressure on the leeward side of the trailer behind the tractor-trailer gap. Furthermore, it is found that the drag reduction performance increases with CVTD length but does not vary with the number of CVTDs between one and four. These results suggest that using a single CVTD or flexible sheet of material at the centerline of the cab-gap is the most viable solution, as there is no further benefit to using multiple devices. In addition, it allows for the greatest CVTD length without impeding articulation