Application of hybrid forecast engine based intelligent algorithm and feature selection for wind signal prediction
- PDF / 1,567,611 Bytes
- 15 Pages / 595.276 x 790.866 pts Page_size
- 3 Downloads / 194 Views
ORIGINAL PAPER
Application of hybrid forecast engine based intelligent algorithm and feature selection for wind signal prediction Mahdi Mir1 · Mahdi Shafieezadeh2 · Mohammad Amin Heidari3 · Noradin Ghadimi4 Received: 24 April 2018 / Accepted: 20 January 2019 © Springer-Verlag GmbH Germany, part of Springer Nature 2019
Abstract This paper presents a new prediction model based on empirical mode decomposition, feature selection and hybrid forecast engine. The whole structure of proposed model is based on nonstationarity and non-convex nature of wind power signal. The hybrid forecast engine consists of three main stages as; empirical mode decomposition, an intelligent algorithm and back propagation neural network. All parameters of proposed neural network will be optimized by intelligent algorithm. Effectiveness of the proposed model is tested with real-world hourly data of wind farms in Spain and Texas. In order to demonstrate the validity of the proposed model, it is compared with several other wind speed and power forecast techniques. Obtained results confirm the validity of the developed approach. Keywords Neural network · Wind power forecast · Hybrid forecast engine · Feature selection · EEMD
1 Introduction A wind power prediction resembles to an evaluation of the predictable generation of one or more wind turbines in the near future. By production is often meant existing power for wind farm measured (Yeh 2013; Hamian et al. 2018). Prediction can also be presented in terms of energy, by integrating power production over each time interval (Leng et al. 2018; Yeh 2017). According to important role of this energy in operation of power system, an accurate prediction model is demanded. So, recently, several prediction models have been proposed. Some of the proposed models can be described as; Auto-Regressive Moving Average (ARMA) models (Yeh et al. 2014; Ahmadian et al. 2014), Fractional ARIMA (FARIMA) model (Abedinia et al. 2017a, b), Auto-Regressive Integrated Moving Average (ARIMA)
* Noradin Ghadimi [email protected] 1
Department of Electrical Engineering, Ferdowsi University of Mashhad, Mashhad, Iran
2
Yazd University, Yazd, Iran
3
Shiraz Electricity Distribution Company (SHEDC), Shiraz, Iran
4
Young Researchers and Elite Club, Ardabil Branch, Islamic Azad University, Ardabil, Iran
model (Mohammadi et al. 2018), NN-based forecast engine in Eskandari Nasab et al. (2014), combination of differential Empirical Mode Decomposition (EMD) and Relevance Vector Machine (RVM) in Abedinia et al. (2017) for prediction of short-term wind power output, two Neural Network (NN) based models in Abedinia and Ghadimi (2013), combination of nonparametric and time-varying regression and time-series model, i.e. Holt–Winters and ARMA in Ghadimi et al. (2018), and Hybrid Iterative Forecast Method (HIFM) in Bagal et al. (2018) based on two stage feature selection model. Although the mentioned approaches are simple and strong forecasting methods and can be easily implemented, most of the predictors are
Data Loading...
