The development of the alternating current field measurement (ACFM) technique for high-speed inspection of railway wheels/track is important to ensure the integrity of railway assets. The stress and conditions at the wheel/rail contact patch, severity of rolling contact fatigue (RCF) damage and changes in microstructure and hardness with tread depth for railway wheels are investigated. This study employs experimental measurement of RCF defects in railway wheels to understand ACFM sensor signal - RCF defect relationships. The influence of sensor frequency and speed, sensor angle relative to crack angle, lift-off distance, crack propagation angle and inter-crack spacing distance are investigated. Low rather than high frequency sensors are better suited at sizing cracks in railway wheels and track. The optimum signal is obtained when the sensor is oriented parallel to the crack angle. The signal is reduced at higher speeds. The maximum change in the normalised Bx signal component of the sensor’s magnetic field is greater for closely spaced cracks (< 5 mm) and thus overestimates the defect. Inter-crack spacing distances of 15-20 mm yield estimates for crack pocket depths that are oversized by 20-36%. The inner cracks in a cluster with four defects yields higher Bₓ values than expected and thus significantly over sizes the defects at 15-20 mm inter-crack spacing distances, whereas, the sensor provides reasonable depth estimates for the outer cracks. Crack propagation angle affects the signal. Scans conducted with the probe oriented at 90° to the cluster of cracks consisting of a deep central crack surrounded by shallow cracks results in a distinct central sensor Bₓ signal that consists of a peak due to a flux leakage effect, thus, identifying the critical defect.

The author of this thesis deals with the issues of real time traffic management in junction areas and bottleneck sections on mainline railways in the event of service disturbances. A systematic methodology is proposed for modelling and solving real time train rescheduling problems in junction areas and bottleneck sections, including train re-sequencing and train re-timing. A formal mathematical model, the Junction Rescheduling Model (JRM) is proposed, based on a Mixed Integer Programming (MIP) to minimise a Weighted Average Delay (WAD). An innovative algorithm based on Differential Evolution algorithm, named DE_JRM is proposed for solving real time train rescheduling problems formulated with JRM. The performance of the algorithm DE_JRM has been evaluated with a stochastic method based on Monte-Carlo simulation methodology. The evaluation results show a good performance for both flyover and flat junctions compared with First Come First Served (FCFS) and a conventional ARS strategy. The author also extends the proposed methodology, including JRM and the algorithm DE_JRM, to model and solve real time train rescheduling problems for bottleneck sections of railway networks. Finally, an integrated system architecture for the traffic management and train control is introduced for system implementation of the proposed methodology of train rescheduling in junction areas and bottleneck sections on mainline railways.

The rail industry has focused on the improvement of maintenance through the effective use of online condition monitoring of rolling stock and rail infrastructure in order to reduce the occurrence of unexpected catastrophic failures and disruption that arises from them to an absolute minimum. The basic components comprising a railway wheelset are the wheels, axle and axle bearings. Detection of wheelset faults in a timely manner increases efficiency as it helps minimise maintenance costs and increase availability. The main aim of this project has been the development of a novel integrated online acoustic emission (AE) and vibration testing technique for the detection of wheel and axle bearing defects as early as possible and well before they result in catastrophic failure and subsequently derailment. The approach employed within this research study has been based on the combined use of accelerometers and high-frequency acoustic emission sensors mounted on the rail or axle box using magnetic hold-downs. Within the framework of this project several experiments have been carried out under laboratory conditions, as well as in the field at the Long Marston Test Track and in Cropredy on the Chiltern Railway line to London.

Through the use of new communication-based train control (CBTC) systems, modern metro railways have been able to provide a more efficient, more reliable and more eco-friendly transport services. The main advantages of the CBTC systems are achieved by utilising modern communication technologies. The performance of the communications network is dependent on a well-designed access point (AP) deployment, as this determines the overall communication capability and impacts the cost. In this thesis, a systematic methodology is proposed for formulating and solving AP deployment planning (ADP) problems in two scenarios: (i) a tunnel section area; and (ii) a real-world metro system. Different mathematical models are presented for modelling the ADP problem in these two scenarios. In addition to mathematical models, an exhaustive search and a customized search algorithm, which uses a multi-objective evolutionary algorithm based on decomposition (MOEA/D), are proposed for solving the ADP optimisation problems. The methodologies are applied to the scenarios mentioned above. To evaluate the optimisation results, the optimised AP deployments are tested on a simulation platform integrating a railway network simulator and a communication network simulator. The test result shows that with the optimised AP deployments the DCS can achieve a better performance while using fewer APs.

Energy consumption in electric locomotives is principally the power consumed in traction motors. In order to reduce this energy consumption, the motion resistances of the train need to be reduced. These resistances include aerodynamics; inertial and grade forces; curving resistance; and bearing and wheel/rail friction. Though many factors such as gradient resistance cannot be modified, if a control system is included, curving resistance can be minimised by reducing the energy losses in the contact patches between wheel and rail. Therefore, operational practices could be modified in order to obtain the most appropriate wheelset attack angle between wheel and rail, and appropriate train speed. One solution is to implement a steering control system. The function of this control system is to monitor and control the wheelset lateral displacement or the attack angle of the wheelset. This could reduce the energy dissipated at the contact points between wheel and rail, consequently reducing the energy consumed by traction motors in railway vehicles. Therefore, the work presented in this thesis aims to design and develop a control method for combined vehicle traction and wheelset active steering control systems and to assess the energy efficiency of a rail vehicle under typical operational conditions. In order to achieve these aims, two dynamic models of a typical railway vehicle have been developed in MATLAB and Simulink. The first model comprises the electrical traction and mechanical system passive system). The second model includes the passive and the wheelset active steering control system (active system). These models are used to determine the relationship between traction energy consumption and the energy dissipated in the contact points between wheel and rail, and to compare the passive steering system to the wheelset active steering control system, determining the possibilities for energy saving. In order to assess the influence of the wheelset active steering control on the relationship between wheel and rail contact forces and traction power a series of deterministic track features are set comprising curve radii with different cant deficiencies and wheel conicities. Also a typical track profile from Leeds to Hull is used. From these simulations, the traction energy consumption, energy dissipated at the contact patches, and energy consumed by the steering actuators are calculated. Statistical analyses are used to understand the relationship between the traction power and wheelset motion dynamics (lateral displacement and attack angle). The active vehicle model scheme is used to investigate the improvement of the energy efficiency of a railway vehicle using active steering. The wheelset active steering control system analysis shows whether different combinations of vehicle speed, wheelset conicity and track curve radius lead to a reduction, no reduction, or an increase intraction power consumption. The probability of high power consumption under different conditions is assessed to ensure that it is reduced wherever possible. The ability of a forecasting model to predict the traction power consumption behaviour of railway vehicles from the wheelset motion dynamic is assessed. Findings show that the overall prediction accuracy is fairly similar to the power measured from the passive vehicle running on a track from Leeds to Hull. However, the algorithm does not perform effectively for the deterministic track features. Finally, the benefits of implementing wheelset active steering control systems in terms of the mitigation of contact forces between wheels and rails and how this mitigation influences traction energy consumption are evaluated to determine under what conditions energy can be saved.

The argument for this thesis is that patterns of past engineering and operational development can be used to support the creation of a strategy for future development and that, in order to achieve this, a corporate understanding of the history of the engineering, operational and organisational changes in the business is essential for any railway undertaking. The research shows that a railway is a complex system, where the operational life of much of the hardware is greater than the time employees spend in service, so it is common for the origins and reasons for systems on the railway to be forgotten. This knowledge can be lost when staff retire or leave. For new staff, little attention is paid to history and development during induction; indeed, such knowledge may not even be considered necessary or relevant. London Underground is used as a case study to show how the history of the system has impacted on the design of rolling stock. The work concludes that it is essential for railway employees to understand the way in which systems and operations have evolved on a railway during its history. The author proposes new processes that include historical corporate knowledge in future business development.

In this thesis, we present a new clustering algorithm we call <em>Significance Feature Clustering</em>, which is designed to cluster text documents. Its central premise is the mapping of raw frequency count vectors to discrete-valued significance vectors which contain values of -1, 0, or 1. These values represent whether a word is <em>significantly negative</em>, <em>neutral</em>, or <em>significantly positive</em>, respectively. Initially, standard tf-idf vectors are computed from raw frequency vectors, then these tf-idf vectors are transformed to significance vectors using a parameter alpha, where alpha controls the mapping -1, 0, or 1 for each vector entry. SFC clusters agglomeratively, with each document's significance vector representing a cluster of size one containing just the document, and iteratively merges the two clusters that exhibit the most similar average using cosine similarity. We show that by using a good alpha value, the significance vectors produced by SFC provide an accurate indication of which words are significant to which documents, as well as the type of significance, and therefore correspondingly yield a good clustering in terms of a well-known definition of clustering quality. We further demonstrate that a user need not manually select an alpha as we develop a new definition of clustering quality that is highly correlated with text clustering quality. Our metric extends the family of metrics known as <em>internal similarity</em>, so that it can be applied to a tree of clusters rather than a set, but it also factors in an aspect of recall that was absent from previous internal similarity metrics. Using this new definition of internal similarity, which we call <em>maximum tree internal similarity</em>, we show that a close to optimal text clustering may be picked from any number of clusterings created by different alpha's. The automatically selected clusterings have qualities that are close to that of a well-known and powerful hierarchical clustering algorithm.

In this thesis, we present a new clustering algorithm we call <em>Significance Feature Clustering</em>, which is designed to cluster text documents. Its central premise is the mapping of raw frequency count vectors to discrete-valued significance vectors which contain values of -1, 0, or 1. These values represent whether a word is <em>significantly negative</em>, <em>neutral</em>, or <em>significantly positive</em>, respectively. Initially, standard tf-idf vectors are computed from raw frequency vectors, then these tf-idf vectors are transformed to significance vectors using a parameter alpha, where alpha controls the mapping -1, 0, or 1 for each vector entry. SFC clusters agglomeratively, with each document's significance vector representing a cluster of size one containing just the document, and iteratively merges the two clusters that exhibit the most similar average using cosine similarity. We show that by using a good alpha value, the significance vectors produced by SFC provide an accurate indication of which words are significant to which documents, as well as the type of significance, and therefore correspondingly yield a good clustering in terms of a well-known definition of clustering quality. We further demonstrate that a user need not manually select an alpha as we develop a new definition of clustering quality that is highly correlated with text clustering quality. Our metric extends the family of metrics known as <em>internal similarity</em>, so that it can be applied to a tree of clusters rather than a set, but it also factors in an aspect of recall that was absent from previous internal similarity metrics. Using this new definition of internal similarity, which we call <em>maximum tree internal similarity</em>, we show that a close to optimal text clustering may be picked from any number of clusterings created by different alpha's. The automatically selected clusterings have qualities that are close to that of a well-known and powerful hierarchical clustering algorithm.

Oι COUP-TFs (Chicken Ovalvumin Upstream Promoter – Transcription Factors) είναι ορφανοί πυρηνικοί υποδοχείς που ανήκουν στην υπεροικογένεια των υποδοχέων στεροειδών/θυρεοειδών ορμονών. Εμφανίζουν μεγάλη ομολογία σε όλα τα μετάζωα και εκφράζονται κατά προτίμηση στο νευρικό σύστημα. Εκτός της συντηρημένης αλληλουχίας τους, επίσης συντηρημένες είναι και οι θέσεις των εσoνίων στους διάφορους οργανισμούς ενώ, μέχρι πρόσφατα, δεν είχε παρατηρηθεί εναλλακτικό μάτισμα. Ωστόσο έχουμε δείξει ότι στα εχινόδερμα, υπάρχουν δύο μετάγραφα, αποτέλεσμα εναλλακτικού ματίσματος ενός μικρού εξονίου 63 bps. Τα δύο mRNAs παράγουν δύο πρωτεΐνες, η μεγαλύτερη εκ των οποίων περιέχει 21 επιπρόσθετα αμινοξέα στην κάρβοξυ-τελική επέκταση της περιοχής πρόσδεσης στο DNA (DBD). Πειράματα EMSA έδειξαν ότι η μικρή ισομορφή του COUP-TF, ως ομοδιμερές, προσδένεται με διαφoρετική συγγένεια σε όλα τα στοιχεία απόκρισης που εξετάστηκαν. Η μεγάλη ισομορφή εν αντιθέσει, δεν προσδένεται αποτελεσματικά σε κανένα από αυτά τα στοιχεία απόκρισης, ούτε ως ομοδιμερές ούτε ως ετεροδιμερές με τη μικρή ισομορφή. Επίσης, πειράματα ανοσοφθορισμού, έχουν δείξει ότι σε έμβρυα αχινού στο στάδιο των 16 κυττάρων, η μεγάλη πρωτεΐνη εντοπίζεται στο κυτταρόπλασμα, ενώ για τη μικρή πρωτεΐνη είναι γνωστή η παρουσία της στον πυρήνα. Ο σκοπός της εργασίας αυτής είναι η μελέτη του ρόλου της ένθεσης των 21 αμινοξέων στην μεγάλη ισομορφή, όσον αφορά στην κυτταρική διαμερισματοποίηση και στη συγγένεια πρόσδεσης στο DNA. Στην ένθεση αυτή υπάρχουν δύο προλίνες και τρείς θρεονίνες. Η προλίνη γενικότερα βρίσκεται σε σημεία όπου μία πρωτεΐνη κάμπτεται. Η υπόθεσή μας είναι ότι οι προλίνες αυτές παίζουν σημαντικό ρόλο στη στερεοδιαμόρφωση της πρωτεΐνης, που με τη σειρά της είναι υπεύθυνη για τις αλλαγές που παρατηρούνται στη μεγάλη ισομορφή, ως προς την συγγένεια πρόσδεσης στο DNA και την κυτταρική διαμερισματοποίηση, σε σχέση με την μικρή. Επίσης η θρεονίνη είναι ένα αμινοξύ που μπορεί να δεκτεί φωσφορυλίωση. Γεγονότα φωσφορυλίωσης και αποφοσφωρυλίωσης μπορεί να επηρεάζουν την ικανότητα πρόσδεσης και την κυτταρική διαμερισματοποίηση του υποδοχέα. 2 Για να ελεχθούν οι υποθέσεις αυτές, μεταλλάχθηκαν οι δύο προλίνες και οι τρεις θρεονίνες της μεγάλης ισομορφής σε αλανίνες, παράχθηκαν in vitro οι μεταλλαγμένες πρωτεΐνες και προσδιορίστηκε η συγγένεια πρόσδεσης τους στο DNA με πειράματα EMSA. Οι ίδιες μεταλλάξεις πραγματοποιήθηκαν και σε ένα κατασκεύασμα το οποίο κωδικοποιεί για τη χιμαιρική πρωτεΐνη που αποτελείται από τη μεγάλη ισομορφή και τη πράσινη φθορίζουσα πρωτεΐνη. Τα in vitro παραχθέντα RNAs ενέθηκαν σε έμβρυα αχινού ώστε να μελετηθεί η ενδοκυτταρική κατανομή της μεγάλης COUP-TF ισομορφής. Τα αποτελέσματα δείχνουν ότι οι δύο προλίνες δεν φαίνεται να παίζουν σημαντικό ρόλο όσον αφορά την πρόσδεση της μεγάλης COUP-TF ισομορφής ως ομοδιμερές. Εντούτοις οι δύο αυτές προλίνες φαίνεται να επηρεάζουν την πρόσδεση της μεγάλης ισομορφής COUP-TF ως ετεροδιμερές με τρόπο ώστε να την ενισχύει. / -

The aim of this thesis is to develop an integrated methodology for investigating the potential for new local railway stations within a given area, with particular emphasis on the use of Geographical Information Systems (GIS). Existing methods for assessing the case for constructing new local railway stations have often been found wanting, with the forecasts produced proving to be inaccurate. A review of previous work in this field has been undertaken and methodologies with the potential to enhance local rail demand models have been identified. Trip rate and trip end models have been developed which are capable of forecasting usage at new station sites anywhere in England and Wales. Geographically Weighted Regression (GWR) has been used to enhance the performance of these models and to account for local variations in the effects of explanatory variables on rail demand. Flow level models have been produced for stations in South-East Wales, with a range of model formulations tested. A survey of ultimate passenger trip origins and destinations was carried out in the same area, enabling the accuracy of theoretical station catchment definition methods to be tested. A GIS-based procedure for locating potential sites for new railway stations within a given area has been developed. This was combined with the results from the demand models and estimates of associated costs and benefits to give a synthesised appraisal procedure capable of assessing the case for constructing particular stations. This procedure was applied to 14 sites in South-East Wales and, along with trip end forecasts for 421 sites across the country, this indicated that there is almost certainly a positive case for constructing a significant number of new railway stations in the UK.