Tuning of DFIG Wind Turbine Controllers with Voltage Regulation Subjected to Electrical Faults Using a PSO Algorithm

• Published in: Journal of control, automation & electrical systems Vol. 32; no. 5; pp. 1417 - 1428
• Main Authors: Aguilar, Milton Ernesto Barrios, Coury, Denis Vinicius, Machado, Fernanda Ribeiro, Reginatto, Romeu
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.10.2021; Springer Nature B.V
• Subjects: Active damping; Algorithms; Control; Control and Systems Theory; Control stability; Controllers; Electric potential; Electrical Engineering; Electrical faults; Engineering; Induction generators; Mechatronics; Parameters; Particle swarm optimization; Proportional integral; Robotics; Robotics and Automation; Stiffness; Tuning; Turbines; Voltage; Wind turbines; Comparison “trial and error”-PSO; Electrical and mechanical modeling; PV controle; Maximum power point tracking; PI controllers
• ISSN: 2195-3880; 2195-3899; 2195-3899
• DOI: 10.1007/s40313-021-00779-w

The doubly fed induction generator (DFIG) wind turbine is the most used in wind generation, and there are different control strategies for this type of machine. The correct choice of the proportional-integral (PI) controller parameters values determine the stability and performance of the DFIG when subjected to a transient disturbances. This research considers the DFIG with active power and voltage (PV) control loops, and seeks to adjust the PI controllers parameters to obtain a damped behavior and small peaks in the electrical and mechanical variables when the wind turbine is subjected to electrical faults. For the tuning, a particle swarm optimization (PSO) algorithm that minimizes an objective function (OF) is proposed, derived from the weighting of three objectives related to the dynamic performance of the wind turbine during the fault. In addition, for comparative purposes another tuning is performed via the “trial and error” method (T&E). The results of computer simulations showed that the PSO is an effective tool for this nonlinear and continuous problem, achieving an adequate tuning of the PI controller parameters, in addition to showing the difficulties of the “trial and error” method, as it is not systematic. The proposed PSO method provided a tuning improving the performance of the penalized quantities, damping the electrical and mechanical oscillations, with low peaks in the active power, terminal voltage remained within the pre-established limits, and preserving performance for different fault intensities and shaft stiffness variations.