Forecasting of 5MW solar photovoltaic power plant generation using generalized neural network
01 Dec 2015-pp 1-6
TL;DR: In this article, a two stage procedure is used referred to as GNN (Generalized Neural Network) model for forecasting the power generated in a 5MW solar PV plant owned by Gujarat Power Corporation Limited (GPCL) at Charanka solar park, Gujarat.
Abstract: . The percentage of renewable energy sources such as solar, wind power and biomass in the energy mix of India is increasing every year. Solar power variability is an important issue for grid integration of solar photovoltaic power plants. The main objective of this paper is to forecast the power generated in a 5 MW solar PV plant owned by Gujarat Power Corporation Limited (GPCL) at Charanka solar park, Gujarat. Charanka is a location with an average of320 sunny days in a year. Average solar insolation available here is 5.7–6.0 kWh/m2 per day. Data obtained from 1st March 2014–31st August 2014 is used for analysis purposes. In this paper a two stage procedure is used referred to as GNN (Generalized Neural Network) model. In the primary stage pre-processing is done on the raw data followed by neural network model for forecasting.
Citations
More filters
23 Nov 2000
TL;DR: The journal provides rapid dissemination of important results in soft computing technologies, a fusion of research in evolutionary algorithms and genetic programming, neural science and neural net systems, fuzzy set theory and fuzzy systems, and chaos theory and chaotic systems.
Abstract: Soft Computing is dedicated to system solutions based on soft computing techniques. It provides rapid dissemination of important results in soft computing technologies, a fusion of research in evolutionary algorithms and genetic programming, neural science and neural net systems, fuzzy set theory and fuzzy systems, and chaos theory and chaotic systems. Soft Computing encourages the integration of soft computing techniques and tools into both everyday and advanced applications. By linking the ideas and techniques of soft computing with other disciplines, the journal serves as a unifying platform that fosters comparisons, extensions, and new applications. As a result, the journal is an international forum for all scientists and engineers engaged in research and development in this fast growing field. Indexed by Science Citation Index, Inspec, Compendex, DBLP, Computer Science Index, Current Abstracts, Current
181 citations
13 Jul 2017
TL;DR: This dissertation proposes the application of Deep Learning frameworks to perform automatic disturbance classification, from which a set of measurements from several PMUs installed in the Low Voltage grid of an interconnected system is used, so as to endow a classifier of knowledge related to system disturbances.
Abstract: The analysis of power system disturbances is fundamental to ensure the reliability and security of the supply. In fact, capturing the sequence of system states over a disturbance is an increased value to understand its origin. Phasor Measurement Units (PMUs) have the ability to record these fast transients with high precision, by providing synchronized measurements at high sampling rates. Indeed, these events can occur in a few seconds, which hampers their detection by the traditional SCADA (Supervisory Control and Data Acquisition) systems and emphasizes the uniqueness of PMUs. With the advent of Wide Area Measurement Systems (WAMS) and the consequent deployment of such monitoring devices, control centers are being flooded with massive volumes of data. Therefore, transforming data into knowledge, preferably automatically, is an actual challenge for system operators. Under abnormal operating conditions, the data collected from several PMUs scattered across the grid can shape a sort of a "movie" of the disturbance. The importance of WAMS is therefore sustained on their ability to capture the sequence of events resulting from a disturbance, helping the further analysis procedures. Driven by the amounts of data involved, this dissertation proposes the application of Deep Learning frameworks to perform automatic disturbance classification. In order to do so, a set of measurements from several PMUs installed in the Low Voltage grid of an interconnected system is used, from which representative patterns are extracted so as to endow a classifier of knowledge related to system disturbances. In particular, the strategies herein adopted consist of the application of Multilayer Perceptrons, Deep Belief Networks and Convolutional Neural Networks, the latter having outperformed the others in terms of classification accuracy. Additionally, these architectures were implemented in both the CPU and the GPU to ascertain the resulting gains in speed.
7 citations
Cites background from "Forecasting of 5MW solar photovolta..."
...Therefore, a lot of subjects have at least experienced their use: for instance, the definition of protection schemes for transmission lines, determining the location of the fault with ANNs [40] [41]; transformer fault diagnosis [42] [43]; on-line voltage stability monitoring [44]; forecasting problems of load [45] [46] [47], wind power [48] [49] and photovoltaic production [50] [51] [52]....
[...]
07 Dec 2020
TL;DR: The authors presente trabalho tem o objetivo de realizar uma previsao de geracao fotovoltaica for a cidade de Garibaldi, Rio Grande do Sul, para um horizonte de seis meses a frente, por meio da modelagem of uma rede neural artificial.
Abstract: presente trabalho tem o objetivo de realizar uma previsao de geracao fotovoltaica para a cidade de Garibaldi, Rio Grande do Sul, para um horizonte de seis meses a frente, por meio da modelagem de uma rede neural artificial. Os dados de entrada desta rede sao compostos por conjuntos de dados historicos climaticos em complemento a um conjunto de dados historicos de geracao solar fotovoltaica que serao usados como alvos. A classificacao das variaveis de entrada e realizada por meio da regressao stepwise, que elenca as variaveis com maior relacao com a resposta. A modelagem conta com o auxilio do software numerico Matlab, para a construcao do algoritmo e a classificacao dos dados. Para o treinamento desta rede neural artificial e empregado o metodo de aprendizado supervisionado deretropropagacao do erro, utilizando a funcao de treinamento bayesian regularization, visando reduzir o erro da previsao. A rede neural artificial com melhor desempenho atingiu um MAPE de 12,97 %. Tambem neste contexto sao apresentadas as comparacoes de desempenho das redes neurais para diferentes horizontes, com resultados alcancados estando de acordo com os encontrados na literatura.
References
More filters
TL;DR: It is rigorously established that standard multilayer feedforward networks with as few as one hidden layer using arbitrary squashing functions are capable of approximating any Borel measurable function from one finite dimensional space to another to any desired degree of accuracy, provided sufficiently many hidden units are available.
18,794 citations
"Forecasting of 5MW solar photovolta..." refers background in this paper
...It has been demonstrated that a multi-layered ANN are universal approximators and are able to approximate any nonlinear continuous function up to the desired level of accuracy [7]....
[...]
TL;DR: In this article, a curve representing a simple harmonic function of the time, and superposing on the ordinates small random errors, is shown to make the graph somewhat irregular, leaving the suggestion of periodicity still quite clear to the eye.
Abstract: If we take a curve representing a simple harmonic function of the time, and superpose on the ordinates small random errors, the only effect is to make the graph somewhat irregular, leaving the suggestion of periodicity still quite clear to the eye Fig 1 ( a ) shows such a curve, the random errors having been determined by the throws of dice If the errors are increased in magnitude, as in fig 1 ( b ), the graph becomes more irregular, the suggestion of periodicity more obscure, and we have only sufficiently to increase the “errors” to mask completely any appearance of periodicity But, however large the errors, periodogram analysis is applicable to such a curve, and, given a sufficient number of periods, should yield a close approximation to the period and amplitude of the underlying harmonic function When periodogram analysis is applied to data respecting any physical phenomenon in the expectation of eliciting one or more true periodicities, there is usually, as it seems to me, a tendency to start from the initial hypothesis that the periodicity or periodicities are masked solely by such more or less random superposed fluctuations — fluctuations which do not in any way disturb the steady course of the underlying periodic function or functions It is true that the periodogram itself will indicate the truth or otherwise of the hypothesis made, but there seems no reason for assuming it to be the hypothesis most likely a priori
1,105 citations
1,019 citations
"Forecasting of 5MW solar photovolta..." refers background in this paper
...In last few years, several authors have proposed regression based conventional solar insolation forecasting models such as Auto regression (AR) and Moving Average (MA) [1, 2]....
[...]
817 citations
"Forecasting of 5MW solar photovolta..." refers methods in this paper
...The use of integrated regression models known as Auto Regressive Moving Average (ARMA) and Auto Regressive integrated Moving Average known as advance models of regression techniques are used for stationary time series [3]....
[...]
TL;DR: An advanced statistical method for solar power forecasting based on artificial intelligence techniques that can be well trained to improve the forecast accuracy and is suitable for operational planning of transmission system operator and for PV power system operators trading in electricity markets.
490 citations
"Forecasting of 5MW solar photovolta..." refers background in this paper
...In [9] weather based neural network model is introduced for operation and planning of solar power transmission....
[...]