Techniques to improve forecasting models: applications to energy demand and price

This thesis is a study of three techniques to improve performance of some standard fore-casting models, application to the energy demand and prices. We focus on forecasting demand and price one-day ahead. First, the wavelet transform was used as a pre-processing procedure with two approaches: multicomponent-forecasts and direct-forecasts. We have empirically compared these approaches and found that the former consistently outperformed the latter. Second, adaptive models were introduced to continuously update model parameters in the testing period by combining ?lters with standard forecasting methods. Among these adaptive models, the adaptive LR-GARCH model was proposed for the fi?rst time in the thesis. Third, with regard to noise distributions of the dependent variables in the forecasting models, we used either Gaussian or Student-t distributions. This thesis proposed a novel algorithm to infer parameters of Student-t noise models. The method is an extension of earlier work for models that are linear in parameters to the non-linear multilayer perceptron. Therefore, the proposed method broadens the range of models that can use a Student-t noise distribution. Because these techniques cannot stand alone, they must be combined with prediction models to improve their performance. We combined these techniques with some standard forecasting models: multilayer perceptron, radial basis functions, linear regression, and linear regression with GARCH. These techniques and forecasting models were applied to two datasets from the UK energy markets: daily electricity demand (which is stationary) and gas forward prices (non-stationary). The results showed that these techniques provided good improvement to prediction performance.

Short-term electricity demand and gas price forecasts using wavelet transforms and adaptive models

This paper presents some forecasting techniques for energy demand and price prediction, one day ahead. These techniques combine wavelet transform (WT) with fixed and adaptive machine learning/time series models (multi-layer perceptron (MLP), radial basis functions, linear regression, or GARCH). To create an adaptive model, we use an extended Kalman filter or particle filter to update the parameters continuously on the test set. The adaptive GARCH model is a new contribution, broadening the applicability of GARCH methods. We empirically compared two approaches of combining the WT with prediction models: multicomponent forecasts and direct forecasts. These techniques are applied to large sets of real data (both stationary and non-stationary) from the UK energy markets, so as to provide comparative results that are statistically stronger than those previously reported. The results showed that the forecasting accuracy is significantly improved by using the WT and adaptive models. The best models on the electricity demand/gas price forecast are the adaptive MLP/GARCH with the multicomponent forecast; their MSEs are 0.02314 and 0.15384 respectively.

The n-tuple network: coping with sub-optimality

The subject of this thesis is the n-tuple net.work (RAMnet). The major advantage of RAMnets is their speed and the simplicity with which they can be implemented in parallel hardware. On the other hand, this method is not a universal approximator and the training procedure does not involve the minimisation of a cost function. Hence RAMnets are potentially sub-optimal. It is important to understand the source of this sub-optimality and to develop the analytical tools that allow us to quantify the generalisation cost of using this model for any given data. We view RAMnets as classifiers and function approximators and try to determine how critical their lack of' universality and optimality is. In order to understand better the inherent. restrictions of the model, we review RAMnets showing their relationship to a number of well established general models such as: Associative Memories, Kamerva's Sparse Distributed Memory, Radial Basis Functions, General Regression Networks and Bayesian Classifiers. We then benchmark binary RAMnet. model against 23 other algorithms using real-world data from the StatLog Project. This large scale experimental study indicates that RAMnets are often capable of delivering results which are competitive with those obtained by more sophisticated, computationally expensive rnodels. The Frequency Weighted version is also benchmarked and shown to perform worse than the binary RAMnet for large values of the tuple size n. We demonstrate that the main issues in the Frequency Weighted RAMnets is adequate probability estimation and propose Good-Turing estimates in place of the more commonly used :Maximum Likelihood estimates. Having established the viability of the method numerically, we focus on providillg an analytical framework that allows us to quantify the generalisation cost of RAMnets for a given datasetL. For the classification network we provide a semi-quantitative argument which is based on the notion of Tuple distance. It gives a good indication of whether the network will fail for the given data. A rigorous Bayesian framework with Gaussian process prior assumptions is given for the regression n-tuple net. We show how to calculate the generalisation cost of this net and verify the results numerically for one dimensional noisy interpolation problems. We conclude that the n-tuple method of classification based on memorisation of random features can be a powerful alternative to slower cost driven models. The speed of the method is at the expense of its optimality. RAMnets will fail for certain datasets but the cases when they do so are relatively easy to determine with the analytical tools we provide.

On the Error-Free Computation of Fast Cosine Transform

We extend our previous work into error-free representations of transform basis functions by presenting a novel error-free encoding scheme for the fast implementation of a Linzer-Feig Fast Cosine Transform (FCT) and its inverse. We discuss an 8x8 L-F scaled Discrete Cosine Transform where the architecture uses a new algebraic integer quantization of the 1-D radix-8 DCT that allows the separable computation of a 2-D DCT without any intermediate number representation conversions. The resulting architecture is very regular and reduces latency by 50% compared to a previous error-free design, with virtually the same hardware cost.

Trained Neural Network Characterizing Variables for Predicting Organic Retention by Nanofiltration Membranes

Many organic compounds cause an irreversible damage to human health and the ecosystem and are present in water resources. Among these hazard substances, phenolic compounds play an important role on the actual contamination. Utilization of membrane technology is increasing exponentially in drinking water production and waste water treatment. The removal of organic compounds by nanofiltration membranes is characterized not only by molecular sieving effects but also by membrane-solute interactions. Influence of the sieving parameters (molecular weight and molecular diameter) and the physicochemical interactions (dissociation constant and molecular hydrophobicity) on the membrane rejection of the organic solutes were studied. The molecular hydrophobicity is expressed as logarithm of octanol-water partition coefficient. This paper proposes a method used that can be used for symbolic knowledge extraction from a trained neural network, once they have been trained with the desired performance and is based on detect the more important variables in problems where exist multicolineality among the input variables.

Development of N N SMRT for N a power of 2 and its Applications in Image Enhancement

Digital Image Processing is a rapidly evolving eld with growing applications in Science and Engineering. It involves changing the nature of an image in order to either improve its pictorial information for human interpretation or render it more suitable for autonomous machine perception. One of the major areas of image processing for human vision applications is image enhancement. The principal goal of image enhancement is to improve visual quality of an image, typically by taking advantage of the response of human visual system. Image enhancement methods are carried out usually in the pixel domain. Transform domain methods can often provide another way to interpret and understand image contents. A suitable transform, thus selected, should have less computational complexity. Sequency ordered arrangement of unique MRT (Mapped Real Transform) coe cients can give rise to an integer-to-integer transform, named Sequency based unique MRT (SMRT), suitable for image processing applications. The development of the SMRT from UMRT (Unique MRT), forward & inverse SMRT algorithms and the basis functions are introduced. A few properties of the SMRT are explored and its scope in lossless text compression is presented.

Symmetry properties of one- and two- electron molecular integrals

The maximum numbers of distinct one- and two-electron integrals that arise in calculating the electronic energy of a molecule are discussed. It is shown that these may be calculated easily using the character table of the symmetry group of the set of basis functions used to express the wave function. Complications arising from complex group representations and from a conflict of symmetry between the basis set and the nuclear configuration are considered and illustrated by examples.

Combustion Instability and Active Control: Alternative Fuels, Augmentors, and Modeling Heat Release

Experimental and analytical studies were conducted to explore thermo-acoustic coupling during the onset of combustion instability in various air-breathing combustor configurations. These include a laboratory-scale 200-kW dump combustor and a 100-kW augmentor featuring a v-gutter flame holder. They were used to simulate main combustion chambers and afterburners in aero engines, respectively. The three primary themes of this work includes: 1) modeling heat release fluctuations for stability analysis, 2) conducting active combustion control with alternative fuels, and 3) demonstrating practical active control for augmentor instability suppression. The phenomenon of combustion instabilities remains an unsolved problem in propulsion engines, mainly because of the difficulty in predicting the fluctuating component of heat release without extensive testing. A hybrid model was developed to describe both the temporal and spatial variations in dynamic heat release, using a separation of variables approach that requires only a limited amount of experimental data. The use of sinusoidal basis functions further reduced the amount of data required. When the mean heat release behavior is known, the only experimental data needed for detailed stability analysis is one instantaneous picture of heat release at the peak pressure phase. This model was successfully tested in the dump combustor experiments, reproducing the correct sign of the overall Rayleigh index as well as the remarkably accurate spatial distribution pattern of fluctuating heat release. Active combustion control was explored for fuel-flexible combustor operation using twelve different jet fuels including bio-synthetic and Fischer-Tropsch types. Analysis done using an actuated spray combustion model revealed that the combustion response times of these fuels were similar. Combined with experimental spray characterizations, this suggested that controller performance should remain effective with various alternative fuels. Active control experiments validated this analysis while demonstrating 50-70\% reduction in the peak spectral amplitude. A new model augmentor was built and tested for combustion dynamics using schlieren and chemiluminescence techniques. Novel active control techniques including pulsed air injection were implemented and the results were compared with the pulsed fuel injection approach. The pulsed injection of secondary air worked just as effectively for suppressing the augmentor instability, setting up the possibility of more efficient actuation strategy.

Discontinuous Galerkin Methods on hp-Anisotropic Meshes I: A Priori Error Analysis

We consider the a priori error analysis of hp-version interior penalty discontinuous Galerkin methods for second-order partial differential equations with nonnegative characteristic form under weak assumptions on the mesh design and the local finite element spaces employed. In particular, we prove a priori hp-error bounds for linear target functionals of the solution, on (possibly) anisotropic computational meshes with anisotropic tensor-product polynomial basis functions. The theoretical results are illustrated by a numerical experiment.

Triangular mesh parameterization with trimmed surfaces

Given a 2manifold triangular mesh \(M \subset {\mathbb {R}}^3\), with border, a parameterization of \(M\) is a FACE or trimmed surface \(F=\{S,L_0,\ldots, L_m\}\) -- \(F\) is a connected subset or region of a parametric surface \(S\), bounded by a set of LOOPs \(L_0,\ldots ,L_m\) such that each \(L_i \subset S\) is a closed 1manifold having no intersection with the other \(L_j\) LOOPs -- The parametric surface \(S\) is a statistical fit of the mesh \(M\) -- \(L_0\) is the outermost LOOP bounding \(F\) and \(L_i\) is the LOOP of the ith hole in \(F\) (if any) -- The problem of parameterizing triangular meshes is relevant for reverse engineering, tool path planning, feature detection, redesign, etc -- Stateofart mesh procedures parameterize a rectangular mesh \(M\) -- To improve such procedures, we report here the implementation of an algorithm which parameterizes meshes \(M\) presenting holes and concavities -- We synthesize a parametric surface \(S \subset {\mathbb {R}}^3\) which approximates a superset of the mesh \(M\) -- Then, we compute a set of LOOPs trimming \(S\), and therefore completing the FACE \(F=\ {S,L_0,\ldots ,L_m\}\) -- Our algorithm gives satisfactory results for \(M\) having low Gaussian curvature (i.e., \(M\) being quasi-developable or developable) -- This assumption is a reasonable one, since \(M\) is the product of manifold segmentation preprocessing -- Our algorithm computes: (1) a manifold learning mapping \(\phi : M \rightarrow U \subset {\mathbb {R}}^2\), (2) an inverse mapping \(S: W \subset {\mathbb {R}}^2 \rightarrow {\mathbb {R}}^3\), with \ (W\) being a rectangular grid containing and surpassing \(U\) -- To compute \(\phi\) we test IsoMap, Laplacian Eigenmaps and Hessian local linear embedding (best results with HLLE) -- For the back mapping (NURBS) \(S\) the crucial step is to find a control polyhedron \(P\), which is an extrapolation of \(M\) -- We calculate \(P\) by extrapolating radial basis functions that interpolate points inside \(\phi (M)\) -- We successfully test our implementation with several datasets presenting concavities, holes, and are extremely nondevelopable -- Ongoing work is being devoted to manifold segmentation which facilitates mesh parameterization

Photovoltaic power forecast modeling with artificial neural networks

Dissertação de Mestrado, Engenharia Eletrónica e Telecomunicações, Faculdade de Ciências e Tecnologia, Universidade do Algarve, 2016

Um modelo matemático baseado em wavelets para análise do método térmico de recuperação de óleo pesado aplicando irradiação eletromagnética

This work proposes a model to investigate the use of a cylindrical antenna used in the thermal method of recovering through electromagnetic radiation of high-viscosity oil. The antenna has a simple geometry, adapted dipole type, and it can be modelled by using Maxwell s equation. The wavelet transforms are used as basis functions and applied in conjunction with the method of moments to obtain the current distribution in the antenna. The electric field, power and temperature distribution are carefully calculated for the analysis of the antenna as electromagnetic heating. The energy performance is analyzed based on thermo-fluid dynamic simulations at field scale, and through the adaptation in the Steam Thermal and Advanced Processes Reservoir Simulator (STARS) by Computer Modelling Group (CMG). The model proposed and the numerical results obtained are stable and presented good agreement with the results reported in the specialized literature

Design Optimization of Nozzle Shapes for Maximum Uniformity of Exit Flow

The objective of this study is to identify the optimal designs of converging-diverging supersonic and hypersonic nozzles that perform at maximum uniformity of thermodynamic and flow-field properties with respect to their average values at the nozzle exit. Since this is a multi-objective design optimization problem, the design variables used are parameters defining the shape of the nozzle. This work presents how variation of such parameters can influence the nozzle exit flow non-uniformities. A Computational Fluid Dynamics (CFD) software package, ANSYS FLUENT, was used to simulate the compressible, viscous gas flow-field in forty nozzle shapes, including the heat transfer analysis. The results of two turbulence models, k-e and k-ω, were computed and compared. With the analysis results obtained, the Response Surface Methodology (RSM) was applied for the purpose of performing a multi-objective optimization. The optimization was performed with ModeFrontier software package using Kriging and Radial Basis Functions (RBF) response surfaces. Final Pareto optimal nozzle shapes were then analyzed with ANSYS FLUENT to confirm the accuracy of the optimization process.

Multi-mode communications in near field exploiting OAM properties

This dissertation analyzes the exploitation of the orbital angular momentum (OAM) of the electromagnetic waves with large intelligent surfaces in the near-field region and line-of-sight conditions, in light of the holographic MIMO communication concept. Firstly, a characterization of the OAM-based communication problem is presented, and the relationship between OAM-carrying waves and communication modes is discussed. Then, practicable strategies for OAM detection using large intelligent surfaces and optimization methods based on beam focusing are proposed. Numerical results characterize the effectiveness of OAM with respect to other strategies, also including the proposed detection and optimization methods. It is shown that OAM waves constitute a particular choice of communication modes, i.e., an alternative basis set, which is sub-optimum with respect to optimal basis functions that can be derived by solving eigenfunction problems. Moreover, even the joint utilization of OAM waves with focusing strategies led to the conclusion that no channel capacity achievements can be obtained with these transmission techniques.