Ricardo A. S. Fernandes

Maximum and minimum stock price forecasting of Brazilian power distribution companies based on artificial neural networks

Graphical abstractDisplay Omitted HighlightsWe predict maximum and minimum day stock prices of power companies.The methodology is based on attribute selection and time series prediction.The most relevant attributes are determined by correlation analysis.The actual time series prediction is carried out by neural networks.The proposed methodology provides very good results. Time series forecasting has been widely used to determine future prices of stocks, and the analysis and modeling of finance time series is an important task for guiding investors decisions and trades. Nonetheless, the prediction of prices by means of a time series is not trivial and it requires a thorough analysis of indexes, variables and other data. In addition, in a dynamic environment such as the stock market, the non-linearity of the time series is a pronounced characteristic, and this immediately affects the efficacy of stock price forecasts. Thus, this paper aims at proposing a methodology that forecasts the maximum and minimum day stock prices of three Brazilian power distribution companies, which are traded in the Sao Paulo Stock Exchange BM&FBovespa. When compared to the other papers already published in the literature, one of the main contributions and novelty of this paper is the forecast of the range of closing prices of Brazilian power distribution companies stocks. As a result of its application, investors may be able to define threshold values for their stock trades. Moreover, such a methodology may be of great interest to home brokers who do not possess ample knowledge to invest in such companies. The proposed methodology is based on the calculation of distinct features to be analysed by means of attribute selection, defining the most relevant attributes to predict the maximum and minimum day stock prices of each company. Then, the actual prediction was carried out by Artificial Neural Networks (ANNs), which had their performances evaluated by means of Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) calculations. The proposed methodology for addressing the problem of prediction of maximum and minimum day stock prices for Brazilian distribution companies is effective. In addition, these results were only possible to be achieved due to the combined use of attribute selection by correlation analysis and ANNs.

Feature Extraction and Power Quality Disturbances Classification Using Smart Meters Signals

This paper presents a methodology aimed at extracting features to obtain information that will highlight disturbances related to the field of power quality. Due to the concept of smart grids, it is clear that the classification of the disturbances should be undertaken using smart meters, so that a large amount of data corresponding to the voltage and current waveforms are not exchanged using the communication channels, i.e., between smart meter and Utilitys database server. Thus, it is necessary to ensure a balance between computational effort (arising from the implementation of algorithms on hardware) and the satisfactory performance of the algorithm for the classification of disturbances. Based on the assumption that the classification task is onerous, this paper proposes a step of feature extraction that may be calculated and analyzed offline using synthetic waveforms/signals, which are subsequently validated using field measurements. It should be noted that this offline analysis is required to determine the most relevant features for the process of classifying each disturbance. However, in order to establish the effectiveness of the feature extraction step, the response of decision trees of the C4.5 type and of artificial neural networks of the multilayer perceptron type were verified during the phase of disturbance classification. In short, good results were obtained that corroborate the hypothesis that the feature extraction step is necessary to classify disturbances effectively and with low computational effort.

Identification of residential load profile in the Smart Grid context

This work presents an automatic method for identification of residential load profile in the Smart Grid context. Hence, in this research were used client/server software interfaces to transmit and receive data through the Internet. In this case, the residential consumers and utility were represented by client and server software, respectively. However, to consider all the stages of this method, a database was created to store fictitious data related to consumer measurements. From these data, the utility software was able to furnish information about consumers load profile and use this information to make decisions. The results were obtained using an experimental workbench that contains residential loads, a configurable power supply and an energy analyzer. It is show in an experimental way some benefits that can be achieved with the introduction of Smart Grid concept on distribution systems.

Adaptive threshold based on wavelet transform applied to the segmentation of single and combined power quality disturbances

Graphical abstractDisplay Omitted HighlightsAn adaptive threshold is determined for segmentation of power quality signals.The power quality signals are based on mathematical models and acquired in field.Wavelet transforms are used to decompose the signals.The intersections between the adaptive threshold and the wavelet transform detail curves determine the start and the end of the segments.The adaptive threshold was accurate in more than 96 percent of the signals. Detecting discontinuities in electrical signals from recorded oscillograms makes it possible to segment them. This is the first step in implementing automated methods which will ensure disturbances in electrical power systems are detected, classified and stored. In this context, this paper presents a way of determining an adaptive threshold based on the decomposition of electrical signals through the Discrete Wavelet Transform (DWT) using Daubechies family filter banks, allowing for the segmentation of signals and, as a consequence, the analysis of disturbances related to Power Quality (PQ). Considering this, the proposed approach was initially evaluated for signals originating from mathematical models representing short-term voltage fluctuations, transients (impulsive and oscillatory) and harmonic distortions. In the synthetic signal database, either single or combined occurrences of more than one disturbance were considered. By applying the DWT, the amount of energy and entropy of energy were then calculated for the leaves of the second level of decomposition. Based on these calculations, a unique adaptive threshold could be determined for each analyzed signal. Afterwards, the amount of existing intersections between the threshold and the curve of details obtained for the second level of decomposition was then defined. Thus, the intersections determine the beginning and end of the segments. In order to validate the approach, the performance of the proposed methodology was analyzed considering the signals obtained from oscillograms provided by IEEE 1159.3 Task Force, as well as real oscillograms obtained from a regional distribution utility. After these analyses, it was observed that the proposed approach is efficient and applicable to automatic segmentation of events related to PQ.

Fault Detection in Power Distribution Systems Using Automated Integration of Computational Intelligence Tools

The main purpose of this paper is to present architecture of automated system that allows monitoring and tracking in real time (online) the possible occurrence of faults and electromagnetic transients observed in primary power distribution networks. Through the interconnection of this automated system to the utility operation center, it will be possible to provide an efficient tool that will assist in decision-making by the Operation Center. In short, the desired purpose aims to have all tools necessary to identify, almost instantaneously, the occurrence of faults and transient disturbances in the primary power distribution system, as well as to determine its respective origin and probable location. The compilations of results from the application of this automated system show that the developed techniques provide accurate results, identifying and locating several occurrences of faults observed in the distribution system.

Identificação de cargas lineares e não-lineares em sistemas elétricos residenciais usando técnicas para seleção de atributos e redes neurais artificiais

Este trabalho consiste em apresentar um metodo para a identificacao de cargas lineares e nao-lineares comumente encontradas em sistemas eletricos residenciais. Desta identificacao, solucoes viaveis poderao ser aplicadas com o intuito de mitigar os niveis de emissao das correntes harmonicas geradas, advindas principalmente por cargas com caracteristicas nao-lineares. No desenvolvimento do metodo, utilizaram-se de tecnicas para a selecao de atributos, de forma a minimizar a dificuldade em se identificar as cargas conectadas ao sistema. A etapa posterior de identificacao foi realizada pela aplicacao de redes neurais artificiais. Todas as situacoes de distorcao harmonica foram geradas em laboratorio por uma fonte de alimentacao, onde em sua saida foram alocados analisadores de energia, responsaveis pela extracao das medidas necessarias sobre as cargas residenciais em analise. Os resultados obtidos foram considerados satisfatorios, mostrando-se que a metodologia proposta pode ser tambem empregada pelas concessionarias de energia eletrica para que estas obtenham informacoes sobre o perfil das cargas instaladas em consumidores residenciais.

MASCEM: EPEX SPOT Day-Ahead market integration and simulation

The energy sector restructuring process in industrialized countries had the aim of reducing electricity prices by increasing competitiveness, and facilitate the integration of distributed energy resources. However, the complexity in market players interactions has increased, and new problems have emerged. Decision support tools that facilitate the study and comprehension of these markets became extremely useful, providing players with competitive advantage. MASCEM (Multi-Agent Simulator of Competitive Electricity Markets) arises in this context, modeling and simulating real electricity markets. It is crucial to MASCEM to have the ability to simulate as many market models and player types as possible, thus enhancing the ability to recreate the electricity markets reality in its maximum possible extent. This paper presents the EPEX Spot Day-Ahead market integration in MASCEM. EPEX Spot SEs mission is to lead European markets coupling in a single unified market, thus being crucial for the study of competitive electricity markets.

Very short-term load forecasting based on NARX recurrent neural networks

Time series forecasting is an important task in various fields of science, like economy, engineering and other areas that use historical data to predict future problems. In this context, Artificial Neural Networks have shown promising results for this task, when compared with the traditional statistical techniques. Thus, this research aims to evaluate the performance of NARX-neural network (Nonlinear Autoregressive Model with Exogenous Input) for the purpose of performing load forecasting for very short-term data from distribution substations. The cross validation was applied to evaluate different topologies. It is important to mention that the data was obtained by measures done in Brazilian substations located at two different cities. The results show the contribution of the paper once it demonstrates the efficiency of the NARX-neural network compared with Feedforward and Elman neural networks, which are widely used to predict times series.

Analysis of Wavelet Transform applied to the segmentation of disturbance signals with different sampling rates

The segmentation of disturbance signals obtained by power quality meters at distribution substations represents a complex task due to the difficult to determine the initial and final points where the disturbance occur. With the implementation of Smart Grids, the segmentation have been a crucial procedure to guarantee the better storage of data and improve the classification of disturbances. In this context, this paper provides a methodology based on Wavelet Transform and the determination of an adaptive threshold that allows the segmentation of voltage and/or current signals. So, we consider in this work short-duration voltage variations, impulsive and oscillatory transients, and harmonic distortions. In order to reach the above mentioned objectives, the disturbance signals were synthetically generated. Moreover, Gaussian noise was convolved on the signals with the intuit to obtain data closely to real power signals (voltage or current). Thus, these signals were used to validate the method proposed.

Power quality disturbances segmentation based on wavelet decomposition and adaptive thresholds

This paper proposes an automatic method for determining an adaptive threshold based on Wavelet decomposition, which allows the segmentation of power quality signals. Primarily, these signals were synthetically generated, which represents short duration voltage variations, oscillatory and impulsive transients and harmonic distortions. Subsequently, the signals were decomposed by Daubechies Wavelets, whose second level of decomposition provided the values of energy and entropy of the energy used in determining the adaptive thresholds. The segments were then obtained by the intersections of the curves of Wavelets details with the calculated thresholds. The results reached with this methodology were evaluated under synthetic and real oscillographs in order to show its efficiency.