A new method for wind speed forecasting based on empirical mode decomposition and improved persistence approach

Abstract

Wind speed forecasting plays an important role in sizing the capacity of the energy storage system and guaranteeing the security and stability of power system. In order to forecast wind speeds more accurately, a hybrid forecasting method based on empirical mode decomposition (EMD) and an improved persistence approach has been proposed in this paper. Employing the EMD technique to decompose the measured wind speeds into many intrinsic mode function (IMF) components and a residue, which represent the original signal in both high-frequency and low-frequency signals. Meanwhile each IMF is analyzed and predicted using Moving Average method (high-frequency signals) and Persistence Approach (low-frequency signals), so does the residue. The sum of the predictive value for each decomposed component is the forecasted data. A set of measured wind speed data from a given wind farm locating at Jiangsu Province in China were modeled using the proposed method and the forecasted results were compared to the measured wind speeds as well as those predicted with other traditional methods. The results indicate that the forecasting precision can be improved with the developed model.

A Review on Hybrid Empirical Mode Decomposition Models for Wind Speed and Wind Power Prediction

“…c. EMD eliminates the noisy patterns, randomness, instability and large fluctuations in the data [12,15,19]. d. Apart from unique signal decomposition, IMFs (generated with EMD) have good local characteristics in both time as well as frequency domains [27]. e. The working principle of EMD is empirical without any mathematical/statistical calculations and hence is very easy to understand [28].…”

“…In Refs. [10,19,27,46] and [60], the IMFs were categorized into two bands with high and low frequency components. These articles stated the IMFs with lower frequency bands represent the central tendency of the data and highly regular pattern which shows the accurate characteristics of the original data, whereas the IMFs with higher frequencies contain large quantity of noisy signals, which mainly reflects the random information that led to a great disturbance for prediction precision of wind data.…”

“…However, Refs. [27,37,60] believed that there is potential and useful information involved in IMF 1 and handled this situation differently. Instead of removing high frequency IMFs, both high and low frequency IMFs were predicted with different and more suitable prediction models.…”

“…Instead of removing high frequency IMFs, both high and low frequency IMFs were predicted with different and more suitable prediction models. In [27], the high frequency IMFs were predicted using a moving average method and the low frequency IMFs with the conventional Persistence approach. Similarly, in [60], the high frequency IMFs were again decomposed and reconstructed by Wavelet transformation.…”

Power Quality Improvement Strategy for Wind Energy Conversion System

“…In the literature, there are several hybrid models consisting EMD and statistical/CI based models such as EMD-SVR [11], EMD-NN [8] and EMD-MA-Persistent [12]. However, there are simpler models consisting EMD and k-Nearest Neighbor (kNN) for TS forecasting.…”