Prediction of postprandial blood glucose under intra-patient variability and uncertainty and its use in the design of insulin dosing strategies for type 1 diabetic patients

In this thesis I propose a novel method to estimate the dose and injection-to-meal time for low-risk intensive insulin therapy. This dosage-aid system uses an optimization algorithm to determine the insulin dose and injection-to-meal time that minimizes the risk of postprandial hyper- and hypoglycaemia in type 1 diabetic patients. To this end, the algorithm applies a methodology that quantifies the risk of experiencing different grades of hypo- or hyperglycaemia in the postprandial state induced by insulin therapy according to an individual patient’s parameters. This methodology is based on modal interval analysis (MIA). Applying MIA, the postprandial glucose level is predicted with consideration of intra-patient variability and other sources of uncertainty. A worst-case approach is then used to calculate the risk index. In this way, a safer prediction of possible hyper- and hypoglycaemic episodes induced by the insulin therapy tested can be calculated in terms of these uncertainties.

Elastic-net prefiltering for two-class classification

A two-stage linear-in-the-parameter model construction algorithm is proposed aimed at noisy two-class classification problems. The purpose of the first stage is to produce a prefiltered signal that is used as the desired output for the second stage which constructs a sparse linear-in-the-parameter classifier. The prefiltering stage is a two-level process aimed at maximizing a model's generalization capability, in which a new elastic-net model identification algorithm using singular value decomposition is employed at the lower level, and then, two regularization parameters are optimized using a particle-swarm-optimization algorithm at the upper level by minimizing the leave-one-out (LOO) misclassification rate. It is shown that the LOO misclassification rate based on the resultant prefiltered signal can be analytically computed without splitting the data set, and the associated computational cost is minimal due to orthogonality. The second stage of sparse classifier construction is based on orthogonal forward regression with the D-optimality algorithm. Extensive simulations of this approach for noisy data sets illustrate the competitiveness of this approach to classification of noisy data problems.

Multi-vehicle planning using RRT-connect

The problem of planning multiple vehicles deals with the design of an effective algorithm that can cause multiple autonomous vehicles on the road to communicate and generate a collaborative optimal travel plan. Our modelling of the problem considers vehicles to vary greatly in terms of both size and speed, which makes it suboptimal to have a faster vehicle follow a slower vehicle or for vehicles to drive with predefined speed lanes. It is essential to have a fast planning algorithm whilst still being probabilistically complete. The Rapidly Exploring Random Trees (RRT) algorithm developed and reported on here uses a problem specific coordination axis, a local optimization algorithm, priority based coordination, and a module for deciding travel speeds. Vehicles are assumed to remain in their current relative position laterally on the road unless otherwise instructed. Experimental results presented here show regular driving behaviours, namely vehicle following, overtaking, and complex obstacle avoidance. The ability to showcase complex behaviours in the absence of speed lanes is characteristic of the solution developed.

Particle swarm optimisation assisted classification using elastic net prefiltering

A novel two-stage construction algorithm for linear-in-the-parameters classifier is proposed, aiming at noisy two-class classification problems. The purpose of the first stage is to produce a prefiltered signal that is used as the desired output for the second stage to construct a sparse linear-in-the-parameters classifier. For the first stage learning of generating the prefiltered signal, a two-level algorithm is introduced to maximise the model's generalisation capability, in which an elastic net model identification algorithm using singular value decomposition is employed at the lower level while the two regularisation parameters are selected by maximising the Bayesian evidence using a particle swarm optimization algorithm. Analysis is provided to demonstrate how “Occam's razor” is embodied in this approach. The second stage of sparse classifier construction is based on an orthogonal forward regression with the D-optimality algorithm. Extensive experimental results demonstrate that the proposed approach is effective and yields competitive results for noisy data sets.

Multi-site evaluation of an urban land-surface model: intra-urban heterogeneity, seasonality and parameter complexity requirements

An extensive off-line evaluation of the Noah/Single Layer Urban Canopy Model (Noah/SLUCM) urban land-surface model is presented using data from 15 sites to assess (1) the ability of the scheme to reproduce the surface energy balance observed in a range of urban environments, including seasonal changes, and (2) the impact of increasing complexity of input parameter information. Model performance is found to be most dependent on representation of vegetated surface area cover; refinement of other parameter values leads to smaller improvements. Model biases in net all-wave radiation and trade-offs between turbulent heat fluxes are highlighted using an optimization algorithm. Here we use the Urban Zones to characterize Energy partitioning (UZE) as the basis to assign default SLUCM parameter values. A methodology (FRAISE) to assign sites (or areas) to one of these categories based on surface characteristics is evaluated. Using three urban sites from the Basel Urban Boundary Layer Experiment (BUBBLE) dataset, an independent evaluation of the model performance with the parameter values representative of each class is performed. The scheme copes well with both seasonal changes in the surface characteristics and intra-urban heterogeneities in energy flux partitioning, with RMSE performance comparable to similar state-of-the-art models for all fluxes, sites and seasons. The potential of the methodology for high-resolution atmospheric modelling application using the Weather Research and Forecasting (WRF) model is highlighted. This analysis supports the recommendations that (1) three classes are appropriate to characterize the urban environment, and (2) that the parameter values identified should be adopted as default values in WRF.

On the application of optimal wavelet filter banks for ECG signal classification

This paper discusses ECG signal classification after parametrizing the ECG waveforms in the wavelet domain. Signal decomposition using perfect reconstruction quadrature mirror filter banks can provide a very parsimonious representation of ECG signals. In the current work, the filter parameters are adjusted by a numerical optimization algorithm in order to minimize a cost function associated to the filter cut-off sharpness. The goal consists of achieving a better compromise between frequency selectivity and time resolution at each decomposition level than standard orthogonal filter banks such as those of the Daubechies and Coiflet families. Our aim is to optimally decompose the signals in the wavelet domain so that they can be subsequently used as inputs for training to a neural network classifier.

Global minimization using an Augmented Lagrangian method with variable lower-level constraints

A novel global optimization method based on an Augmented Lagrangian framework is introduced for continuous constrained nonlinear optimization problems. At each outer iteration k the method requires the epsilon(k)-global minimization of the Augmented Lagrangian with simple constraints, where epsilon(k) -> epsilon. Global convergence to an epsilon-global minimizer of the original problem is proved. The subproblems are solved using the alpha BB method. Numerical experiments are presented.

Algoritmos para o problema de localização de uma facilidade com distâncias limitadas e restrições de atendimento

The objective in the facility location problem with limited distances is to minimize the sum of distance functions from the facility to the customers, but with a limit on each distance, after which the corresponding function becomes constant. The problem has applications in situations where the service provided by the facility is insensitive after a given threshold distance (eg. fire station location). In this work, we propose a global optimization algorithm for the case in which there are lower and upper limits on the numbers of customers that can be served

Maximização da penetração da geração distribuída através do algoritmo de otimização nuvem de partículas

This work develops a methodology for defining the maximum active power being injected into predefined nodes in the studied distribution networks, considering the possibility of multiple accesses of generating units. The definition of these maximum values is obtained from an optimization study, in which further losses should not exceed those of the base case, i.e., without the presence of distributed generation. The restrictions on the loading of the branches and voltages of the system are respected. To face the problem it is proposed an algorithm, which is based on the numerical method called particle swarm optimization, applied to the study of AC conventional load flow and optimal load flow for maximizing the penetration of distributed generation. Alternatively, the Newton-Raphson method was incorporated to resolution of the load flow. The computer program is performed with the SCILAB software. The proposed algorithm is tested with the data from the IEEE network with 14 nodes and from another network, this one from the Rio Grande do Norte State, at a high voltage (69 kV), with 25 nodes. The algorithm defines allowed values of nominal active power of distributed generation, in percentage terms relative to the demand of the network, from reference values

Avaliacao de dados de secagem de suspensoes de polpas de frutas em leito de jorro com alimentacao intermitente

The objective of this work was the development and improvement of the mathematical models based on mass and heat balances, representing the drying transient process fruit pulp in spouted bed dryer with intermittent feeding. Mass and energy balance for drying, represented by a system of differential equations, were developed in Fortran language and adapted to the condition of intermittent feeding and mass accumulation. Were used the DASSL routine (Differential Algebraic System Solver) for solving the differential equation system and used a heuristic optimization algorithm in parameter estimation, the Particle Swarm algorithm. From the experimental data food drying, the differential models were used to determine the quantity of water and the drying air temperature at the exit of a spouted bed and accumulated mass of powder in the dryer. The models were validated using the experimental data of drying whose operating conditions, air temperature, flow rate and time intermittency, varied within the limits studied. In reviewing the results predicted, it was found that these models represent the experimental data of the kinetics of production and accumulation of powder and humidity and air temperature at the outlet of the dryer

Planning and Projects of Secondary Electric Power Distribution Systems

This paper proposes a methodology to achieve integrated planning and projects for secondary distribution circuits. The planning model is formulated as a mixed integer nonlinear programming problem (MINLP). In order to resolve this problem, a tabu search (TS) algorithm is used, with a neighborhood structure developed to explore the physical characteristics of specific geographies included in the planning and expansion of secondary networks, thus obtaining effective solutions as well as low operating costs and investments. The project stage of secondary circuits consists of calculating the mechanical efforts to determine the support structures of the primary and secondary distribution systems and determining the types of structures that should be used in the system according to topological and electrical parameters of the network and, therefore, accurately assessing the costs involved in the construction and/or reform of secondary systems. A constructive heuristic based on information of the electrical and topological conditions between the medium voltage and low voltage systems is used to connect the primary systems and secondary circuits. The results obtained from planning and design simulations of a real secondary system of electric energy distribution are presented.

Aplicação da técnica de imunofluorescência direta para o diagnóstico da leishmaniose visceral canina em aspirado de linfonodo

Recentemente, foco de leishmaniose visceral canina (CVL) foi descrito na região noroeste do Estado de São Paulo - Brasil. O Hospital Veterinário - UNESP - Araçatuba, no ano de 2.000, desenvolveu 60 testes citopatológicos de casos suspeitos de leishmaniose usando aspirado por agulha fina (FNA). Os esfregaços de linfonodo foram corados pelo método de Romanowsky (Diff-Quik®) e observados em microscopia de luz. Os casos positivos mostraram formas amastigotas típicas de Leishmania livres ou em vacúolos de macrófagos. Sinais citopatológicos de reatividade do sistema linfo-histiocitário com ausência de parasitos foram também observados. Com o objetivo de implementar o diagnóstico da CVL, detectando parasitos e material antigênico nos esfregaços, aplicou-se a reação de imunofluorescência direta (IFD) usando anticorpo policlonal anti-Leishmania produzido em camundongo. Comparamos o método de IFD com a pesquisa direta do parasito em esfregaços corados pelo método de Romanowsky. Dos 60 cães com sinais clínicos da doença, o exame direto foi positivo em 50% (n=30), duvidoso em 36,7% (n=22) e negativo com reatividade do linfonodo em 13,3% (n=8). Quando os linfonodos foram submetidos a reação de IFD observamos reação positiva em 93,3% (n=56) e reação negativa em 6,7% (n=4). Nossos resultados mostraram que a reação de IFD apresentou alta sensibilidade quando comparada a pesquisa direta do parasito pela coloração de Romanowsky. A reação de IFD pode ser um método útil para confirmar os casos duvidosos da doença, onde as formas amastigotas não são identificadas com facilidade.