A high number of instruments that assess various quality characteristics of interest that have an inherent variability monitors hydroelectric plants. The readings of these instruments generate time series of data on many occasions have correlation. Each project of a dam plant has characteristics that make it unique. Faced with the need to establish statistical control limits for the instrumentation data, this article makes an approach to multivariate statistical analysis and proposes a model that uses principal components control charts and statistical T 2 and to explain variability and establish a method of monitoring to control future observations. An application for section E of the Itaipu hydroelectric plant is performed to validate the model. The results show that the method used is appropriate and can help identify the type of outliers, reducing false alarms and reveal instruments that have higher contribution to the variability.
Este artigo propõe a combinação linear das previsões obtidas por três métodos de previsão (a saber, ARIMA, Amortecimento Exponencial e Redes Neurais Artificiais) cujos pesos adaptativos determinados por meio de um problema de programação não linear multiobjetivo, em que se busca minimizar, simultaneamente, as estatísticas: MAE, MAPE e MSE. Os resultados alcançados pela combinação proposta são comparados com a abordagem tradicional de combinação linear de previsões, onde os pesos adaptativos ótimos são determinados somente pela minimização do MSE; com o método de combinação por média aritmética; e com os métodos individuais.
The information provided by accurate forecasts of solar energy time series are considered essential for performing an appropriate prediction of the electrical power that will be available in an electric system, as pointed out in Zhou et al. (2011). However, since the underlying data are highly non-stationary, it follows that to produce their accurate predictions is a very difficult assignment. In order to accomplish it, this paper proposes an iterative Combination of Wavelet Artificial Neural Networks (CWANN) which is aimed to produce shortterm solar radiation time series forecasting. Basically, the CWANN method can be split into three stages: at first one, a decomposition of level p, defined in terms of a wavelet basis, of a given solar radiation time series is performed, generating Wavelet Components (WC); at second one, these WCs are individually modeled by the k different ANNs, where , and the 5 best forecasts of each WC are combined by means of another ANN,
A previsão de séries temporais é largamente utilizada nas diversas áreas do conhecimento humano, principalmente no planejamento e direcionamento estratégico das empresas. O sucesso desta tarefa depende das técnicas de previsões aplicadas. Neste artigo, é proposta uma metodologia híbrida para se projetar séries temporais. Para a validação da metodologia foi escolhida uma série de tempo já modelada por outros autores, possibilitando a comparação dos resultados. A metodologia proposta integra as seguintes técnicas: encolhimento wavelet, decomposição wavelet de nível r e redes neurais artificiais (RNAs). Primeiramente, uma série temporal a ser prevista é submetida ao método de filtragem wavelet proposto, o qual a decompõe em componentes de tendência e de resíduo linear. Em seguida, ambas são decompostas via decomposição de nível r, gerando, para cada uma, r+1 componentes wavelet (CWs); e, em seguida, cada CW é individualmente modelada por uma RNA. Finalmente, as previsões para todas as CWs são linearmente combinadas, produzindo as previsões para a série temporal supracitada. Para avaliá-lo, a série temporal de Canadian Lynx foi usada e todos os resultados alcançados pelo método proposto foram melhores do que outros existentes na literatura.
To improve time series forecasts the wavelet decomposition has been applied. The combination of forecasting methods as the Autoregressive Integrated Moving Average (ARIMA) and Artificial Neural Networks have been used to achieve a higher quality time series forecasting than. This paper proposed a hybrid model composed of wavelet decomposition, ARIMA and neural network Multilayer Perceptron. These models are combined linearly then yielding the time series forecasting. The series studied are the Wolf's sunspots and the British pound/US dollar exchange rate data. The comparison of the proposed model in this paper with literature indicated an effective way to improve forecasting.
7432Eliete Nascimento Pereira et al.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.