Search

The possibility to manipulate the purely noise-induced behaviour in the large ensemble of globally coupled excitable systems is central to my research work. We employ globally coupled noise-driven FitzHugh-Nagumo units as a prototype of excitable system, which serve as a rough model of a neural network. Such a network is capable of demonstrating various kinds of behaviour with non-synchronized or synchronized units, with the mean field demonstrating periodic or chaotic small oscillations, or periodic or aperiodic spiking. Delayed feedback control applied through the mean field is shown capable of manipulating the basic features of the network behaviour, namely, to induce or suppress collective synchrony, to regularize the system behaviour in both synchronous and non-syncrhonous states, to shift the basic time scales of oscillations. These results are relevant to the control of unwanted behaviour in neural networks.

Making an accurate estimation of peak water demand in buildings is essential for engineers and designers in order to ensure proper sizing of water supply systems, storage tanks, boilers, and booster pumps. Over recent years, the amount of potable water used in buildings has reduced considerably as a result of the prevalence of water-efficient appliances and a heightened awareness of the need to conserve water. This has, in part, led to oversizing of water supply networks; a phenomenon that has given cause for concern to those responsible for the design of building plumbing systems. This oversizing problem does not only result in a material and financial cost, it also has negative health consequences. In the UK, despite a clear reduction in consumption at end-use points, traditional design approaches are still used for determining design flow rate. Although different design methods have been presented in various British standards and guidance documents, all use the Loading Unit (LU) approach, which is based on the application of probabilistic techniques, to estimate the design flow for both residential and non-residential buildings. In recent studies, the focus has generally been on residential buildings and there has been little, if any, research to assess the validity of current design methods for non-residential buildings. This study, therefore, focuses on developing a new design approach to estimate demand flow in non-residential buildings. This research starts by providing background information on the water situation in the UK and discusses the reasons for oversizing and its consequences. Water demand is also discussed, as is water conservation, per capita water consumption and the demand from micro-components. In addition, the history of system design and the most commonly used UK design approaches are discussed. After undertaking a critical review and comprehensive investigation of statistical methods and recent studies used to estimate demand flow, a new design methodology for estimating water demand, specifically for non-residential buildings, is introduced. This has also allowed for the presentation of a new stochastic model, namely the Water Demand Estimation Model (WDEM). The model is underpinned by the interaction between users and the provision of sanitary appliances in conjunction with the generation of a comprehensive range of probabilities to capture all possible simultaneous uses of appliances. The Monte Carlo technique has been applied to calculate flow rate values based on a given number of users. A specific type of non-residential building i.e. the 'workplace' has been selected for application of the model and for which new design equations have been derived. Taking into account the water saving appliances used in modern plumbing systems, five design equations have been derived based on efficiency levels of corresponding appliances. In order to validate the model and to assess its accuracy, high quality flow rate data was gathered from three case study buildings. The effectiveness of the WDEM and its impact on the oversizing of water systems has been confirmed by comparing simulated, measured and design flow rates. The results show that the simulated demand is very close to the measured flow rate, and that its use results in a significant reduction of design flow rate compared to those determined by using current design codes. The main outcome of this study is hence a novel approach for the estimation of demand flow for non-residential buildings and a set of design equations to estimate the simultaneous demand flow rate for workplaces. This new approach will be of value to all engineers and designers who seek to establish a more accurate estimation of water demand in buildings.

Described as the most essential natural resource, rivers rank amongst those ecosystems most sensitive to climate change. The 2018 Brisbane Declaration highlights the pressing need to consider the resultant hydroecological impact. To this end, this thesis looks to develop a coupled hydrological-hydroecological modelling framework, an exciting first step under the new research agenda. Initially, the focus lies on developing current understanding of the hydroecological relationship through consideration of potential delays in hydroecological response, alongside refinement of current modelling practice. There follows consideration of whether hydrological models can preserve ecologically relevant characteristics of the flow regime, as determined through hydroecological modelling efforts. Limiting factors are identified and an alternative hydrological modelling approach established. A holistic depiction of uncertainty is central to all developments. The framework is developed with reference to a principal case study, the groundwater-fed River Nar, Norfolk; validation of the component models is achieved through additional case-studies. The hydrological model, forced with climate change projections, is used to simulate changes in the flow regime. This output then serves as input to the coupled hydroecological model. It is thus possible to assess the impact of climate change on hydroecological response in a quantitative manner. Given data limitations, the framework is best suited to applications at the regional scale or by flow regime type. Its importance lies in the potential to inform water resources adaptation, as well as advancing the fields of hydroecological and hydrological modelling. Scope for further research centres around the wider socio-economic context, as recommended under the Brisbane Declaration.

Considering the computational effort and expertise required to simulate 2D hydrodynamic models, it is widely understood that it is practically impossible to run these types of models during a real-time flood event. To allow for real-time flood forecasting and mapping, an automated, computationally efficient and robust data driven modelling engine - as an alternative to the traditional 2D hydraulic models - has been proposed. The concept of computationally efficient model relies heavily on replacing time consuming 2D hydrodynamic software packages with a simplified model structure that is fast, reliable and can robustly retains sufficient accuracy for applications in real-time flood forecasting, mapping and sequential updating. This thesis presents a novel data-driven modelling framework that uses rainfall data from meteorological stations to forecast flood inundation maps. The proposed framework takes advantage of the highly efficient machine learning (ML) algorithms and also utilities the state-of-the-art hydraulic models as a system component. The aim of this research has been to develop an integrated system, where a data-driven rainfall-streamflow forecasting model sets up the upstream boundary conditions for the machine learning based classifiers, which then maps out multi-step ahead flood extents during an extreme flood event. To achieve the aim and objectives of this research, firstly, a comprehensive investigation was undertaken to search for a robust ML-based multi-step ahead rainfall-streamflow forecasting model. Three potential models were tested (Support Vector Regression (SVR), Deep Belief Network (DBN) and Wavelet decomposed Artificial Neural Network (WANN)). The analysis revealed that SVR-based models perform most efficiently in forecasting streamflow for shorter lead time. This study also tested the portability of model parameters and performance deterioration rates. Secondly, multiple ML-based models (SVR, Random Forest (RF) and Multi-layer Perceptron (MLP)) were deployed to simulate flood inundation extents. These models were trained and tested for two geomorphologically distinct case study areas. In the first case of study, of the models trained using the outputs from LISFLOOD-FP hydraulic model and upstream flow data for a large rural catchment (Niger Inland Delta, Mali). For the second case of study similar approach was adopted, though 2D Flood Modeller software package was used to generate target data for the machine learning algorithms and to model inundation extent for a semi-urban floodplain (Upton-Upon-Severn, UK). In both cases, machine learning algorithms performed comparatively in simulating seasonal and event based fluvial flooding. Finally, a framework was developed to generate flood extent maps from rainfall data using the knowledge learned from the case studies. The research activity focused on the town of Upton-Upon-Severn and the analysis time frame covers the flooding event of October-November 2000. RF-based models were trained to forecast the upstream boundary conditions, which were systematically fed into MLP-based classifiers. The classifiers detected states (wet/dry) of the randomly selected locations within a floodplain at every time step (e.g. one hour in this study). The forecasted states of the sampled locations were then spatially interpolated using regression kriging method to produce high resolution probabilistic inundation (9m) maps. Results show that the proposed data centric modelling engine can efficiently emulate the outcomes of the hydraulic model with considerably high accuracy, measured in terms of flood arrival time error, and classification accuracy during flood growing, peak, and receding periods. The key feature of the proposed modelling framework is that, it can substantially reduce computational time, i.e. ~14 seconds for generating flood maps for a flood plain of ~4 km2 at 9m spatial resolution (which is significantly low compared to a fully 2D hydrodynamic model run time).