Augmenting effectiveness of control loops of a PMSG (permanent magnet synchronous generator) based wind energy conversion system by a virtually adaptive PI (proportional integral) controller

Abstract

Offering substantial features, PMSG (permanent magnet synchronous generator) based WECS (wind energy conversion system) is definitely one of the most reliable and efficient ways of extracting electrical power from the wind. Like other WECSs, PMSG-based WECS (PMSG_WECS) encompasses two main control loops, each equipped with PI (proportional integral) controller, to control speed and currents of the system. This work develops a virtually adaptive PI controller to enhance the performance of both main control loops of a PMSG_WECS. A WNN (wavelet neural network) is proposed to be added to each closed control loop in series with PI controller. Due to having a cascade connection, the transfer function of the WNN, which is a pure gain in each time step, is multiplied by PI gains. Therefore, the value of transfer function of the WNN, and consequently, both parameters of PI controller can be changed in each time step by online training of the WNN, resulting in a virtually adaptive PI controller. The performance of the proposed controller in improving efficacy of both current and speed control loops is evaluated by simulation studies and is also compared to that of PI controller, WNNC (wavelet neural network controller), and QNNC (quantum neural network controller).