Abstract:
Decision-making in the selection of technologies for resource recovery from wastewater is challenging due to the many potential recoverable resources and available technologies, with further complexity arising from the multiple objectives of legislative, social, and commercial concerns. To address this the NEREUS Project has developed a decision support tool (DST) to assist decision makers (policy makers, water authorities, process engineers) in identifying the most appropriate technologies for the recovery of a range of resources and scenarios. In this talk, we present the framework and interface of the DST, the interaction between the inputs, parameters, and the multi-criteria scenarios used to derive combinations of unit processes and demonstrates the evidence-based outcomes for resource recovery decisions. The DST provides sustainable solutions generated using multi-criteria optimization considering technical, economic, environmental, and social dimensions as selection criteria. The environmental impact of the suggested solutions is considered via an integrated life cycle analysis integrated into DST. The DST developed in this study enables a transparent, consistent, and informative evaluation of the generated treatment trains to inform and facilitate the implementation of resource recovery technologies.
Abstract: The electrochemistry and electrical behaviour of a Lithium-ion battery cell are typically modelled by the Doyle-Fuller-Newman (DFN) model, which is also known as Porous Electrode Theory or the Pseudo-2D model. The model is comprised of a system of nonlinear partial differential equations that must, in general, be solved numerically. The computational costs associated with solving the DFN model, even for a single planar cell, are relatively high; for example, most commercial solvers take several minutes to simulate a single discharge. This makes its application to computationally expensive problems (such as cell optimisation, parameter estimation, or battery pack simulation) problematic. Motivated by the lack of a fast DFN solver in the literature we have developed a new software tool, named DandeLiion (Dande-Li-ion). This tool provides extremely fast code which solves the DFN model in the cloud within seconds and produces the output of the simulation results directly in the user’s web browser. In order to illustrate the power of DandeLiion we perform a thermally coupled simulation in a realistic pouch cell geometry and in an entire module of battery cells including liquid cooling. We note that such thermally coupled pack simulations are normally performed using simple equivalent circuit models of the cell, rather than the physics-based DFN model, because the computational cost associated with a thermally heterogenous DFN model has, up until now, been too high.
Abstract
From medieval times to the modern day, female mathematicians, real and fictional, have been represented in a variety of ways, both in pictures and in words. By studying such representations, we learn about the women represented and about the attitudes prevailing at the time the representations were created. But deeper questions are invited. How did these representations shape the types of mathematical knowledge women were able to claim? What effect do these representations have on modern-day viewers and readers? Do they, for instance, continue to marginalise the mathematical expertise of women? How can they be used to encourage the participation of women in the mathematical community today?
Abstract: In this talk, I will outline some of the "innovation" projects I have been involved in at the Institute of Cosmology and Gravitation. These are projects that take techniques from astronomy/astrophysics and apply them to other areas of research. The project I will focus on most is research into the spread of diseases using fluorescent dye and code designed for telescopes. We have done several experiments using this new technique to visualise droplet from coughs and sneezes, allowing us to understand how diseases spread and how to better advise cleaning practicing. This is especially relevant since the pandemic began, but is just one example of the surprising work we are doing at the ICG, and I will mention several other projects too.
Abstract: Single photon sensing cameras provide an opportunity for new forms of quantum imaging and sensing. I will overview our work in quantum imaging and sensing using SPAD arrays, with a specific focus on experimental methods for detecting photon pairs and the resulting forms of imaging, e.g. quantum holography and Hong-Ou-Mandel imaging.
Abstract: The complexity is a function on shape space and foliates it by isocomplexity surfaces. The extrema of the complexity are remarkably interesting. The shapes corresponding to them have properties much like those that the Pauli exclusion principle enforces in quantum mechanics. Moreover, a measure is defined intrinsically within any isocomplexity surface. It gives a Born-type probability for shapes of any given complexity, which can be identified with time. Without any use of a wave function, this parallels energy-eigenvalue quantum predictions for configurations of subsystems of the universe. There are several other striking effects associated with shapes that have a given value of the complexity including correlations between widely separated points that might provide a natural explanation of entanglement.
Abstract:
The threat from Space debris is well known having received much publicity. Predominantly the focus for debris surveillance is on Low Earth Orbit (LEO) objects for which radar can be used for tracking objects down to a few centimetres. Were as for Geostationary Earth Orbits (GEO) at an altitude in the region of 40,000 km only optical surveillance is possibly with typically metre sized telescopes. The GEO region of space is a protected region with a fixed number of so-called location slots, which is virtually full. The GEO region is a strategically important region of space for the UK providing vital world-wide communication for defence and civilian use for example. While there are in the region of 160 one metre sized telescopes around the world dedicated to the surveillance of GEO objects these concentrate on monitoring active satellites and known debris down to approximately 0.4 m in size. Over the last few years several GEO satellites have been destroyed by unknown objects. In addition, several spent rocket bodies in Highly Elliptical Orbits (HEO) which cross the GEO protected zone have disintegrated by causes unknown. At a total cost of approximately 1 billion pounds a GEO satellite is an expensive national resource. A team led by Warwick University in 2021 used 8 nights of the 2.54 m Isaac Newton Telescope on La Palma to search 111 sq. degrees i.e. 1 % of the GEO protected zone and found 226 previously unknown objects, 129 of which they could identify as having GEO tracks. Single photon spectrally resolving detectors in the form of Microwave Inductance Detectors (MKIDs) offer the prospect of significantly reducing the size of the telescope and also pushing the observing limit down to potential objects of a few centimetres at GEO. The significant reduction in telescope size offers a significant cost reduction and thus the prospect of a dedicated survey system to understand the debris density in the GEO protected region. While previously the subject of science fiction, the active investigation by the UK government and the European Space agency in space-based solar power as a source of green energy necessitates the understanding of the GEO environment. My talk will discuss the anticipated debris at GEO, the effects of the space environment on satellites, MKIDs and our (Hepburn and Zurvan) work in modelling an MKID debris survey systems and the data we anticipate.
Abstract: In this talk, we study the problem of minimizing a nonconvex quadratic function over a polytope defined by box constraints. This fundamental problem arises in various applications and also as a subproblem in some general nonlinear programming algorithms. Despite its very special structure, this is an NP-hard problem. In fact, it is NP-hard to even approximate a local minimizer. On the other hand, the problem admits various tractable convex relaxations that yield lower bounds on the optimal value. A convex relaxation is said to be exact if the lower bound is equal to the optimal value. We focus on the well-known RLT and RLT-SDP relaxations of box-constrained quadratic programs (BoxQPs). We present complete descriptions of the set of instances that admit exact RLT and RLT-SDP relaxations. We show that our descriptions can be converted into algorithms for efficiently constructing instances of BoxQPs with exact and/or inexact relaxations.
Abstract:
Platelets supply chain (PLTs SC) management is always a challenging task for the healthcare system due to the nature of platelets (PLTs). PLTs have an extremely short shelf life after being extracted from a human body and the demand is highly uncertain, which may lead to a high percentage of wastage and shortage in the PLTs SC. The purpose of this study is to investigate the opportunity of incorporating frozen PLTs (FPLTs) into the PLTs SC to see how it can improve the performance of the PLTs SC in which the PLTs are used only in liquid form with respect to the platelet shortage, wastage and substitution in transfusions. This paper proposes a two-stage stochastic programming (2SSP) model for the PLTs SC with FPLTs and compares it with the model without FPLTs. In order to see how the actual goals are achieved for the PLTs SC with FPLTs, an extended goal programming model is built based on the proposed 2SSP model. In addition, we generate scenarios based on the real data provided by the healthcare practitioners using the combination of a top-down forecasting approach and a Monte Carlo based scenario generation method. After scenario generation, in-sample and out-of-sample stability tests are used to derive a reliable number of scenarios to use in the 2SSP model. By comparing the PLTs SC model with and without FPLTs, the PLTs SC model with FPLTs is shown to be superior to the PLTs model without FPLTs. Furthermore, the output of the PLTs SC model incorporating FPLTs provides managerial insight for the practitioners.
Abstract:
Li-ion batteries (LIBs) are expected to be one of the key technologies used to store electrical energy over the coming years. These batteries are already produced in large quantities due to their high energy storage, high cell voltage, and slow charge reduction when not in use. Their popularity is expected to continue growing with the increasing demand for electric vehicles (EVs) and other devices. Therefore it is important to also focus on improving their performance in terms of their life, safety, and capacity. In comparison to experiments in battery research, models are often used because they offer a cost-effective, safe, and reproducible way to study battery performance, as well as a powerful tool for understanding, optimizing, and predicting the behaviour of new designs. The Doyle Fuller-Newman (DFN) model is the ubiquitous physics-based model for LIBs and is made up of complicated non-linear partial differential equations. It is computationally expensive. Besides, the single particle model (SPM) is a simplified mathematical model which is much faster than DFN model. However, the SPM is based on simplifying assumptions and these assumptions may not accurately reflect the complex physics that occur within a real battery, leading to inaccurate predictions of battery behaviour. This motivates us to systematically simplify a DFN-based model using asymptotic approximations; thereby yielding a reduced-order model that we term the corrected single particle model (cSPM). The cSPM is significantly cheaper to solve than the DFN, yet is able to accurately replicate much of its behaviour. In this presentation we will outline our asymptotic simplification, compare simulation results from cSPM to the DFN and discuss some practical cases where we expect the cSPM to be useful.
Abstract: The glycemic response to a glucose stimulus is an essential tool for detecting deficiencies in humans such as diabetes. In the presence of constant and periodic glucose infusions in healthy individuals, it is known that this control leads to slow oscillations as a result of feedback mechanisms at the organ and tissue level. These ultradian oscillations are typically modelled using systems of nonlinear equations with two discrete delays and here we give a particular attention to its periodic solutions. These arise from a Hopf bifurcation which is induced by an external glucose stimulus and the joint contributions of delays in pancreatic insulin release and hepatic glycogenesis. The effect of each physiological subsystem on the amplitude and period of the oscillations is exhibited by performing a perturbative analysis of its periodic solutions. It is shown that assuming the commensurateness of delays enables the Hopf bifurcation curve to be characterised by studying roots of linear combinations of Chebyshev polynomials. The impact of periodic (sinusoidal) infusions is characterised through numerical bifurcation analysis. The resulting expressions provide an invaluable tool for studying the interplay between physiological functions and delays in producing an oscillatory regime, as well as relevant information for glycemic control strategies.
Joint work with Maia Angelova, Gemma Kirkham and Stefan Ruschel
Abstract:
Geodesics are the "straight lines" of curved spaces, and are a key notion in geometry with lots of applications. If we imagine a spray of geodesics emanating from a single point, then due to the curvature of the space these geodesics may come to focus, much like light passing through a lens or the bright line on the top of your coffee cup. The set of focal points of a spray of geodesics is known as the "conjugate locus", which can be a very complex object. In this talk we will describe novel ways to visualize the conjugate locus and describe a new coordinate system specially adapted to it, from which a greater understanding follows, based on work Matt Cherrie did during his PhD. The presentation will be mostly talk-and-chalk with as little jargon as possible.
class PDF(object):
def __init__(self, pdf, size=(200,200)):
self.pdf = pdf
self.size = size
def _repr_html_(self):
return '<iframe src={0} width={1[0]} height={1[1]}></iframe>'.format(self.pdf, self.size)
PDF('JVabstract.pdf',size=(900,800))
Abstract: A recently developed approach to statistical thermodynamics shows that many paradigmatic mean-field models can be formulated in terms of c-integrable conservation laws of hydrodynamic type with prescribed initial conditions. Examples are the van der Waals model for isotropic fluids, the Curie-Weiss model for magnetism and generalised multi-partite spin systems. The occurrence of phase transitions in such models is explained naturally in terms of the breaking mechanism of nonlinear wave solutions to hyperbolic conservation laws, with the consequent emergence and propagation of classical (dissipative) shock waves. The talk aims at discussing this approach by analysing the discrete Maier-Saupe model for nematic liquid crystals with external fields and novel generalisations for so-called biaxial nematics via the study of a 4-component hydrodynamic-type system. The work presented is based on collaborations with Antonio Moro (Northumbria), Giulio Landolfi (Università del Salento) and Giovanni De Matteis (Università del Salento).
Abstract: We consider evolutionary quasilinear systems of first order u = V(u) u . In the case that V(u) is a diagonal matrix (in particular, with distinct eigenvalues), there is a huge literature investigating the integrability, the solutions (by means of the generalized hodograph method) and the Hamiltonian structures with first order homogeneous Hamiltonian operators. Some recent developments have been achieved by E.V. Ferapontov, M.V. Pavlov and L. Xue in the case when the matrix V(u) is block-diagonal with several upper-triangular blocks. We refer to such systems as being of Jordan block (Toeplitz block) type. In the present talk, we wonder when a quasilinear system in Jordan block form is Hamiltonian, with first order homogeneous operators. Surprisingly, we show that for such systems linear degeneracy is a necessary condition to be Hamiltonian. Finally, we investigate the Hamiltonian structure of El’s kinetic equations for soliton gases after delta-functional reductions. We find that the above mentioned structure is obtained for separable 2-soliton interaction kernels. This talk is based on a joint work with E.V. Ferapontov.