OPTIMAL DESIGN OF A SMART WIND TURBINE BASED ON NEURAL NETWORK, EVOLUTIONARY OPTIMIZATION TECHNIQUE AND CFD

J.L. MROGINSKI; J.M. PODESTÁ; H.G. CASTRO

New York

Conferencia; 13th World Congress on Computational Mechanics (WCCM XIII), 2nd Pan American Congress on Computational Me- chanics (PANACM II); 2018

IACM

It is well known that one of the global priorities is to implement new technologies that allow us to use in a more efficient manner alternative and renewable energy sources, in order to reduce the consumption of conventional energy sources. Wind is one of the most promising renewable energy and, for a number of years, several different types of wind turbines were used, mainly classified as horizontal axis wind turbines (HAWT?s) and vertical axis wind turbines (VAWT?s). Nevertheless, there are still many issues to explore or improve in this subject. Regarding its type, some researches support the fact that VAWT?s perform better in urban environments than HAWT?s [1]. In this work a new model of VAWT is proposed, designed as a combination of the Savonius and Darrieus VAWT?s, where the simplest model is the so-called ?H? turbine. The optimization of the proposed model is made by coupling computational fluid dynamics (CFD) and neural networks, leading to a ?smart? wind turbine with the capability to switch between Savonius and Darrieus types according to the velocity and direction of the incident wind. Furthermore, it is shown how changing length and angle of attack of the blades modify its performance for different wind speeds. The prototype consists of three blades of NACA 0018 profile, initially vertical, with 2 meters high (H) and radius (R) of 1 meter. By means of an opening mechanical device, the operational principle can be interchanged, allowing to start the turbine at relatively low wind speeds. For the CFD simulation, the open-source toolbox OpenFOAM is used with sliding mesh on the rotational domain and SST k-w turbulence model [2]. Considering the wind incidence angle together with the angle of attack of the blades and the average radius of the turbine as optimization variables, the optimal performance of the wind turbine is obtained associated to the maximum energy production and equipment regularity. [1] Li, Q; Maeda, T.; Kamada, Y.; Hiromori, Y.; Nakai, A.; Kasuya, T.; ?Study on stall behavior of a straight-bladed vertical axis wind turbine with numerical and experimental investigations? Journal of Wind Engineering and Industrial Aerodynamics, 164 pp.1-12, 2017. [2] H.G. Castro, R.R. Paz, J.L. Mroginski, M.A. Storti "Evaluation of the proper coherence representation in random flow generation based methods" Journal of Wind Engineering and Industrial Aerodynamics, 168, pp 211?227 (2017)