Robust automatic mapping algorithms in a network monitoring scenario

Automatically generating maps of a measured variable of interest can be problematic. In this work we focus on the monitoring network context where observations are collected and reported by a network of sensors, and are then transformed into interpolated maps for use in decision making. Using traditional geostatistical methods, estimating the covariance structure of data collected in an emergency situation can be difficult. Variogram determination, whether by method-of-moment estimators or by maximum likelihood, is very sensitive to extreme values. Even when a monitoring network is in a routine mode of operation, sensors can sporadically malfunction and report extreme values. If this extreme data destabilises the model, causing the covariance structure of the observed data to be incorrectly estimated, the generated maps will be of little value, and the uncertainty estimates in particular will be misleading. Marchant and Lark [2007] propose a REML estimator for the covariance, which is shown to work on small data sets with a manual selection of the damping parameter in the robust likelihood. We show how this can be extended to allow treatment of large data sets together with an automated approach to all parameter estimation. The projected process kriging framework of Ingram et al. [2007] is extended to allow the use of robust likelihood functions, including the two component Gaussian and the Huber function. We show how our algorithm is further refined to reduce the computational complexity while at the same time minimising any loss of information. To show the benefits of this method, we use data collected from radiation monitoring networks across Europe. We compare our results to those obtained from traditional kriging methodologies and include comparisons with Box-Cox transformations of the data. We discuss the issue of whether to treat or ignore extreme values, making the distinction between the robust methods which ignore outliers and transformation methods which treat them as part of the (transformed) process. Using a case study, based on an extreme radiological events over a large area, we show how radiation data collected from monitoring networks can be analysed automatically and then used to generate reliable maps to inform decision making. We show the limitations of the methods and discuss potential extensions to remedy these.

Fast algorithms for automatic mapping with space-limited covariance functions

In this paper we discuss a fast Bayesian extension to kriging algorithms which has been used successfully for fast, automatic mapping in emergency conditions in the Spatial Interpolation Comparison 2004 (SIC2004) exercise. The application of kriging to automatic mapping raises several issues such as robustness, scalability, speed and parameter estimation. Various ad-hoc solutions have been proposed and used extensively but they lack a sound theoretical basis. In this paper we show how observations can be projected onto a representative subset of the data, without losing significant information. This allows the complexity of the algorithm to grow as O(n m 2), where n is the total number of observations and m is the size of the subset of the observations retained for prediction. The main contribution of this paper is to further extend this projective method through the application of space-limited covariance functions, which can be used as an alternative to the commonly used covariance models. In many real world applications the correlation between observations essentially vanishes beyond a certain separation distance. Thus it makes sense to use a covariance model that encompasses this belief since this leads to sparse covariance matrices for which optimised sparse matrix techniques can be used. In the presence of extreme values we show that space-limited covariance functions offer an additional benefit, they maintain the smoothness locally but at the same time lead to a more robust, and compact, global model. We show the performance of this technique coupled with the sparse extension to the kriging algorithm on synthetic data and outline a number of computational benefits such an approach brings. To test the relevance to automatic mapping we apply the method to the data used in a recent comparison of interpolation techniques (SIC2004) to map the levels of background ambient gamma radiation. © Springer-Verlag 2007.

Hidden state models for time series

Amongst all the objectives in the study of time series, uncovering the dynamic law of its generation is probably the most important. When the underlying dynamics are not available, time series modelling consists of developing a model which best explains a sequence of observations. In this thesis, we consider hidden space models for analysing and describing time series. We first provide an introduction to the principal concepts of hidden state models and draw an analogy between hidden Markov models and state space models. Central ideas such as hidden state inference or parameter estimation are reviewed in detail. A key part of multivariate time series analysis is identifying the delay between different variables. We present a novel approach for time delay estimating in a non-stationary environment. The technique makes use of hidden Markov models and we demonstrate its application for estimating a crucial parameter in the oil industry. We then focus on hybrid models that we call dynamical local models. These models combine and generalise hidden Markov models and state space models. Probabilistic inference is unfortunately computationally intractable and we show how to make use of variational techniques for approximating the posterior distribution over the hidden state variables. Experimental simulations on synthetic and real-world data demonstrate the application of dynamical local models for segmenting a time series into regimes and providing predictive distributions.

An advanced real-time predictive maintenance framework for large scale machine systems

This thesis introduces and develops a novel real-time predictive maintenance system to estimate the machine system parameters using the motion current signature. Recently, motion current signature analysis has been addressed as an alternative to the use of sensors for monitoring internal faults of a motor. A maintenance system based upon the analysis of motion current signature avoids the need for the implementation and maintenance of expensive motion sensing technology. By developing nonlinear dynamical analysis for motion current signature, the research described in this thesis implements a novel real-time predictive maintenance system for current and future manufacturing machine systems. A crucial concept underpinning this project is that the motion current signature contains infor­mation relating to the machine system parameters and that this information can be extracted using nonlinear mapping techniques, such as neural networks. Towards this end, a proof of con­cept procedure is performed, which substantiates this concept. A simulation model, TuneLearn, is developed to simulate the large amount of training data required by the neural network ap­proach. Statistical validation and verification of the model is performed to ascertain confidence in the simulated motion current signature. Validation experiment concludes that, although, the simulation model generates a good macro-dynamical mapping of the motion current signature, it fails to accurately map the micro-dynamical structure due to the lack of knowledge regarding performance of higher order and nonlinear factors, such as backlash and compliance. Failure of the simulation model to determine the micro-dynamical structure suggests the pres­ence of nonlinearity in the motion current signature. This motivated us to perform surrogate data testing for nonlinearity in the motion current signature. Results confirm the presence of nonlinearity in the motion current signature, thereby, motivating the use of nonlinear tech­niques for further analysis. Outcomes of the experiment show that nonlinear noise reduction combined with the linear reverse algorithm offers precise machine system parameter estimation using the motion current signature for the implementation of the real-time predictive maintenance system. Finally, a linear reverse algorithm, BJEST, is developed and applied to the motion current signature to estimate the machine system parameters.

The application of H controller synthesis to high speed independent drive systems

This thesis describes work completed on the application of H controller synthesis to the design of controllers for single axis high speed independent drive design examples. H controller synthesis was used in a single controller format and in a self-tuning regulator, a type of adaptive controller. Three types of industrial design examples were attempted using H controller synthesis, both in simulation and on a Drives Test Facility at Aston University. The results were benchmarked against a Proportional, Integral and Derivative (PID) with velocity feedforward controller (VFF), the industrial standard for this application. An analysis of the differences between a H and PID with VFF controller was completed. A direct-form H controller was determined for a limited class of weighting function and plants which shows the relationship between the weighting function, nominal plant and the controller parameters. The direct-form controller was utilised in two ways. Firstly it allowed the production of simple guidelines for the industrial design of H controllers. Secondly it was used as the controller modifier in a self-tuning regulator (STR). The STR had a controller modification time (including nominal model parameter estimation) of 8ms. A Set-Point Gain Scheduling (SPGS) controller was developed and applied to an industrial design example. The applicability of each control strategy, PID with VFF, H, SPGS and STR, was investigated and a set of general guidelines for their use was determined. All controllers developed were implemented using standard industrial equipment.

Hydrodynamics and mass transfer studies in a scheibel extractor

A 10 cm diameter four-stage Scheibel column with dispersed phase wetted packing sections has been constructed to study the hydrodynamics and mass transfer using the system toluene-acetone-water. The literature pertaining to the above extractor has been examined and the important phenomena such as droplet break-up and coalescence, mass transfer and backmixing have been reviewed. A critical analysis of the backmixing or axial mixing models and the corresponding techniques for parameter estimation was applied and an optimization technique based on Marquardt's algorithm was implemented. A single phase sampling technique was developed to estimate the acetone concentration profile in both phases along the column. Column flooding characteristics were investigated under various operating conditions and it was found that, when the impellers were located at about DI/5cm from the upper surface of the pads, the limiting flow rates increased with impeller speed. This unusual behaviour was explained in terms of the pumping effect created by the turbine impellers. Correlations were developed to predict Sauter mean drop diameters. A five-cell with backflow model was used to estimate the column performance (stage efficiency) and phases non-ideality (backflow parameters). Overall mass transfer coefficients were computed using the above model and compared with those calculated using the correlations based on single drop mechanism.

Optimal design for correlated processes with input-dependent noise

Optimal design for parameter estimation in Gaussian process regression models with input-dependent noise is examined. The motivation stems from the area of computer experiments, where computationally demanding simulators are approximated using Gaussian process emulators to act as statistical surrogates. In the case of stochastic simulators, which produce a random output for a given set of model inputs, repeated evaluations are useful, supporting the use of replicate observations in the experimental design. The findings are also applicable to the wider context of experimental design for Gaussian process regression and kriging. Designs are proposed with the aim of minimising the variance of the Gaussian process parameter estimates. A heteroscedastic Gaussian process model is presented which allows for an experimental design technique based on an extension of Fisher information to heteroscedastic models. It is empirically shown that the error of the approximation of the parameter variance by the inverse of the Fisher information is reduced as the number of replicated points is increased. Through a series of simulation experiments on both synthetic data and a systems biology stochastic simulator, optimal designs with replicate observations are shown to outperform space-filling designs both with and without replicate observations. Guidance is provided on best practice for optimal experimental design for stochastic response models. © 2013 Elsevier Inc. All rights reserved.

Reduced waiting times using a fast switching dual-polarization DDQPSK receiver in a packet switched network

In this paper, we demonstrate a fast switching dual polarization DDQPSK packet switched receiver with very short waiting times. The system employs mth power DDQPSK decoding for high frequency offset tolerance, and Stokes parameter estimation for robust polarization demultiplexing.

Variational mean-field algorithm for efficient inference in large systems of stochastic differential equations

This work introduces a Gaussian variational mean-field approximation for inference in dynamical systems which can be modeled by ordinary stochastic differential equations. This new approach allows one to express the variational free energy as a functional of the marginal moments of the approximating Gaussian process. A restriction of the moment equations to piecewise polynomial functions, over time, dramatically reduces the complexity of approximate inference for stochastic differential equation models and makes it comparable to that of discrete time hidden Markov models. The algorithm is demonstrated on state and parameter estimation for nonlinear problems with up to 1000 dimensional state vectors and compares the results empirically with various well-known inference methodologies.

Multiresponse Robust Engineering: Case with Errors in Factor Levels

2000 Mathematics Subject Classification: 62J05, 62J10, 62F35, 62H12, 62P30.

On Supra-Bayesian Weighted Combination of Available Data Determined by Kerridge Inaccuracy and Entropy

2010 Mathematics Subject Classification: 94A17.

Modelagem e simulação da solubilidade de sais em sistemas aquosos com Monoetilenoglicol

Primary processing of natural gas platforms as Mexilhão Field (PMXL-1 ) in the Santos Basin, where monoethylene glycol (MEG) has been used to inhibit the formation of hydrates, present operational problems caused by salt scale in the recovery unit of MEG. Bibliographic search and data analysis of salt solubility in mixed solvents, namely water and MEG, indicate that experimental reports are available to a relatively restricted number of ionic species present in the produced water, such as NaCl and KCl. The aim of this study was to develop a method for calculating of salt solubilities in mixed solvent mixtures, in explantion, NaCl or KCl in aqueous mixtures of MEG. The method of calculating extend the Pitzer model, with the approach Lorimer, for aqueous systems containing a salt and another solvent (MEG). Python language in the Integrated Development Environment (IDE) Eclipse was used in the creation of the computational applications. The results indicate the feasibility of the proposed calculation method for a systematic series of salt (NaCl or KCl) solubility data in aqueous mixtures of MEG at various temperatures. Moreover, the application of the developed tool in Python has proven to be suitable for parameter estimation and simulation purposes

Recuperação e purificação do antígeno 503 de Leishmania i. chagasi expresso em E. coli e remoção de endotoxina utilizando adsorção em leito expandido

The growing interest and applications of biotechnology products have increased the development of new processes for recovery and purification of proteins. The expanded bed adsorption (EBA) has emerged as a promising technique for this purpose. It combines into one operation the steps of clarification, concentration and purification of the target molecule. Hence, the method reduces the time and the cost of operation. In this context, this thesis aim was to evaluate the recovery and purification of 503 antigen of Leishmania i. chagasi expressed in E. coli M15 and endotoxin removal by EBA. In the first step of this study, batch experiments were carried out using two experimental designs to define the optimal adsorption and elution conditions of 503 antigen onto Streamline chelating resin. For adsorption assays, using expanded bed, it was used a column of 2.6 cm in diameter by 30.0 cm in height coupled to a peristaltic pump. In the second step of study, the removal of endotoxin during antigen recovery process was evaluated employing the non-ionic surfactant Triton X-114 in the washing step ALE. In the third step, we sought developing a mathematical model able to predict the 503 antigen breakthrough curves in expanded mode. The experimental design results to adsorption showed the pH 8.0 and the NaCl concentration of 2.4 M as the optimum adsorption condition. In the second design, the only significant factor for elution was the concentration of imidazole, which was taken at 600 mM. The adsorption isotherm of the 503 antigen showed a good fit to the Langmuir model (R = 0.98) and values for qmax (maximum adsorption capacity) and Kd (equilibrium constant) estimated were 1.95 mg/g and 0.34 mg/mL, respectively. Purification tests directly from unclarified feedstock showed a recovery of 59.2% of the target protein and a purification factor of 6.0. The addition of the non-ionic surfactant Triton X-114 to the washing step of EBA led to high levels (> 99%) of LPS removal initially present in the samples for all conditions tested. The mathematical model obtained to describe the 503 antigen breakthrough curves in Streamline Chelanting resin in expanded mode showed a good fit for both parameter estimation and validation steps. The validated model was used to optimize the efficiencies, achieving maximum values of the process and of the column efficiencies of 89.2% and 75.9%, respectively. Therefore, EBA is an efficient alternative for the recovery of the target protein and removal of endotoxin from an E. coli unclarified feedstock in just one step.

An automatic method of above-water iceberg 3-D profiling based on autonomous surface craft (ASC) using surface images and LIDAR

This thesis reports on a novel method to build a 3-D model of the above-water portion of icebergs using surface imaging. The goal is to work towards the automation of iceberg surveys, allowing an Autonomous Surface Craft (ASC) to acquire shape and size information. After collecting data and images, the core software algorithm is made up of three parts: occluding contour finding, volume intersection, and parameter estimation. A software module is designed that could be used on the ASC to perform automatic and fast processing of above-water surface image data to determine iceberg shape and size measurement and determination. The resolution of the method is calculated using data from the iceberg database of the Program of Energy Research and Development (PERD). The method was investigated using data from field trials conducted through the summer of 2014 by surveying 8 icebergs during 3 expeditions. The results were analyzed to determine iceberg characteristics. Limitations of this method are addressed including its accuracy. Surface imaging system and LIDAR system are developed to profile the above-water iceberg in 2015.

Tsunami simulation and detection using global navigation satellite system reflectometry

In this thesis, research for tsunami remote sensing using the Global Navigation Satellite System-Reflectometry (GNSS-R) delay-Doppler maps (DDMs) is presented. Firstly, a process for simulating GNSS-R DDMs of a tsunami-dominated sea sur- face is described. In this method, the bistatic scattering Zavorotny-Voronovich (Z-V) model, the sea surface mean square slope model of Cox and Munk, and the tsunami- induced wind perturbation model are employed. The feasibility of the Cox and Munk model under a tsunami scenario is examined by comparing the Cox and Munk model- based scattering coefficient with the Jason-1 measurement. A good consistency be- tween these two results is obtained with a correlation coefficient of 0.93. After con- firming the applicability of the Cox and Munk model for a tsunami-dominated sea, this work provides the simulations of the scattering coefficient distribution and the corresponding DDMs of a fixed region of interest before and during the tsunami. Fur- thermore, by subtracting the simulation results that are free of tsunami from those with presence of tsunami, the tsunami-induced variations in scattering coefficients and DDMs can be clearly observed. Secondly, a scheme to detect tsunamis and estimate tsunami parameters from such tsunami-dominant sea surface DDMs is developed. As a first step, a procedure to de- termine tsunami-induced sea surface height anomalies (SSHAs) from DDMs is demon- strated and a tsunami detection precept is proposed. Subsequently, the tsunami parameters (wave amplitude, direction and speed of propagation, wavelength, and the tsunami source location) are estimated based upon the detected tsunami-induced SSHAs. In application, the sea surface scattering coefficients are unambiguously re- trieved by employing the spatial integration approach (SIA) and the dual-antenna technique. Next, the effective wind speed distribution can be restored from the scat- tering coefficients. Assuming all DDMs are of a tsunami-dominated sea surface, the tsunami-induced SSHAs can be derived with the knowledge of background wind speed distribution. In addition, the SSHA distribution resulting from the tsunami-free DDM (which is supposed to be zero) is considered as an error map introduced during the overall retrieving stage and is utilized to mitigate such errors from influencing sub- sequent SSHA results. In particular, a tsunami detection procedure is conducted to judge the SSHAs to be truly tsunami-induced or not through a fitting process, which makes it possible to decrease the false alarm. After this step, tsunami parameter estimation is proceeded based upon the fitted results in the former tsunami detec- tion procedure. Moreover, an additional method is proposed for estimating tsunami propagation velocity and is believed to be more desirable in real-world scenarios. The above-mentioned tsunami-dominated sea surface DDM simulation, tsunami detection precept and parameter estimation have been tested with simulated data based on the 2004 Sumatra-Andaman tsunami event.