Fuzzy-based approach for power smoothing of a full-converter wind turbine generator using a supercapacitor energy storage

Highlights

• The fuzzy approach was able to improve wind power smoothing.

• The proposed approach securely maintained the availability of the supercapacitor.

• The supercapacitor has protected the DC-link of the wind turbine generator.

• The LVRT requirements were fulfilled.

• The supercapacitor’s state of charge was managed after the disturbance.

Abstract

Wind turbine generators (WTGs) are one of the fastest growing renewable energy source technologies. Due to the nature of wind, power fluctuations of WTGs can cause significant problems in the distribution network. In this study a fuzzy-based approach is proposed for a full-converter WTG coupled with a supercapacitor energy storage system. The fuzzy system is designed to smooth out the wind power fluctuations and also maintain an energy reserve of the supercapacitor for short-term grid disturbances. The fuzzy approach is thoroughly tested and compared with a conventional power smoothing technique and with the case without an energy storage system (ESS). Closed-loop digital simulations showed that the proposed fuzzy scheme enhances wind power smoothing and properly manages the state of charge (SOC) of the supercapacitor during faults in the simulated microgrid. Having less power fluctuations and the availability of the ESS to comply with the low-voltage ride through (LVRT) requirement, the WTG and the microgrid operation were considerably improved.

Introduction

Considerable developments in power electronics, turbines and generators have made wind power one of the fastest growing renewable energy source technologies [1]. In fact, wind energy is the most cost-effective renewable source and its share in distribution systems and grid-tied AC microgrids have become significant in recent years [2]. However, due to the stochastic characteristic of the wind, distributed wind turbine generators (WTGs)s may increase the complexity of the microgrid operation. Sudden wind speed variations and the consequent power fluctuations are the source of several technical issues that must be addressed to increase the grid’s reliability. Power fluctuations in weaker grids can produce serious deviations in the voltage and frequency and even cause protective relay devices to trip [3], [4].

The substantial progress in energy storage, as well as the cost reduction of power converters, have made energy storage systems (ESSs)s a feasible solution to improve power quality, efficiency and reliability in networks with significant penetration of renewable sources [5], [6], [7]. ESSss can smooth out wind power fluctuations, regulate short-term voltage and frequency deviations and enhance the low-voltage ride through (LVRT) capability of WTGss [7]. Amongst a variety of ESSs for high-power short-term applications, the supercapacitor is a promising technology and has already shown its effectiveness in power smoothing [8], [9]. However, in order to present the desired performance, a non-trivial control of the ESSs power and the simultaneous management of its state of charge (SOC) is required.

Many studies have presented techniques using the ESSs for power smoothing considering wind power generation. In [8] a fuzzy-based strategy is proposed for the supercapacitor to improve the operation of the doubly-fed induction generator (DFIG) wind turbine. References [10] and [2] also have used a fuzzy system to smooth wind power fluctuations by means of a flywheel ESSs. However, in [10], a flywheel was connected to the DC-link of the DFIG wind turbine while in [2] the ESSs was connected to the grid in parallel with the WTGss. In [6] the authors tested a power smoothing and LVRT control of the supercapacitor for improving the autonomous grid operation. An optimal energy management strategy of a flywheel storage for power smoothing of a full-converter WTGs was proposed in [11].

Although these studies have shown practical and innovative approaches, they have not compared their proposed methods with power smoothing based on conventional techniques and most of them also did not use quantitative indicators to assess the power smoothing performance. Several wind power smoothing methods were tested and numerically compared in [12], both with and without energy storage. In [5] different power smoothing methods based on the first order low-pass filter (FOLPF) were tested and compared through a numerical power fluctuation index. Another approach based on the FOLPF with an adaptive algorithm to self-tune proportional-integral (PI) controllers of the ESSs was proposed in [13] and compared with an approach based on genetic algorithm. A recurrent fuzzy neural network was proposed in [14] for wind power smoothing and was compared with other smoothing methods according to the power and energy capacity required of a battery to maintain the fluctuations within a certain limit. In [15], the moving average filter considering wind power forecast is compared with the conventional moving average technique for wind power smoothing. Nevertheless, these papers have not investigated the SOC management of the ESSs for short-term grid disturbances, such as faults and LVRT.

This paper proposes an integrated approach based on a fuzzy inference system to improve the full-converter WTGs operation. The fuzzy system is designed to continuously perform enhanced power smoothing and also to manage the SOC of the supercapacitor to maintain the ESSs available for short-term grid disturbances. The proposed approach is thoroughly compared with a conventional power smoothing technique using different turbulent wind profiles and suitable quantitative indicators. The fuzzy technique is also tested under a faulted grid condition and compared with the typical WTGs configuration where the ESSs is not considered. The supercapacitor ESSs, used for short-term applications in this study, is coupled in the DC-link of the back-to-back converter of a 2 MW full-converter WTGs. A microgrid test system based on the medium voltage CIGRE benchmark network is used for the analysis. The fuzzy-based approach is implemented in the RSCAD software and tested in the real-time digital simulator (RTDS®).

This paper is organized as follows. Section 2 presents the full-converter WTGs configuration with and without the ESSs, as well as the supercapacitor model. Details of the proposed fuzzy-based and the conventional power smoothing technique are given in Section 3. The test microgrid is presented in Section 4. The simulation procedures to assess the proposed fuzzy-based approach and the results are given in Section 5, which is followed by the conclusions of the paper.