Optimizing the wind power generation in low wind speed areas using an advanced hybrid RBF neural network coupled with the HGA-GSA optimization method

• Published in: Journal of mechanical science and technology Vol. 30; no. 10; pp. 4735 - 4745
• Main Authors: Assareh, Ehsanolah, Poultangari, Iman, Tandis, Emad, Nedaei, Mojtaba
• Format: Journal Article
• Language: English
• Published: Seoul Korean Society of Mechanical Engineers 01.10.2016; Springer Nature B.V; 대한기계학회
• Subjects: Adjustment; Alternative energy sources; Control; Dynamical Systems; Electric power generation; Energy industry; Engineering; Generators; Genetic algorithms; Gravitation; Industrial and Production Engineering; Investigations; Mechanical Engineering; Neural networks; Optimization; Search algorithms; Torque; Vibration; Wind power; Wind power generation; Wind speed; Wind turbines; Windpowered generators; 기계공학; HGA-GSA; Wind generator; Torque regulation; Radial basis function (RBF) neural network; PI controller
• ISSN: 1738-494X; 1976-3824
• DOI: 10.1007/s12206-016-0945-4

Enhancing the energy production from wind power in low-wind areas has always been a fundamental subject of research in the field of wind energy industry. In the first phase of this research, an initial investigation was performed to evaluate the potential of wind in south west of Iran. The initial results indicate that the wind potential in the studied location is not sufficient enough and therefore the investigated region is identified as a low wind speed area. In the second part of this study, an advanced optimization model was presented to regulate the torque in the wind generators. For this primary purpose, the torque of wind turbine is adjusted using a Proportional and integral (PI) control system so that at lower speeds of the wind, the power generated by generator is enhanced significantly. The proposed model uses the RBF neural network to adjust the net obtained gains of the PI controller for the purpose of acquiring the utmost electricity which is produced through the generator. Furthermore, in order to edify and instruct the neural network, the optimal data set is obtained by a Hybrid genetic algorithm along with a gravitational search algorithm (HGA-GSA). The proposed method is evaluated by using a 5MW wind turbine manufactured by National Renewable Energy Laboratory (NREL). Final results of this study are indicative of the satisfactory and successful performance of the proposed investigated model.