A DSP‐based implementation of fuzzy logic and predictive current control for a Sheppard–Taylor power factor correction converter

Abstract

Ensuring a high electric power quality of power systems and electric machinery continues to be a major challenge for power engineers. The research on harmonic reduction and improvement of power factor in connection to alternating current (AC) networks has intensified. This paper focuses on the simulation and real‐time implementation of a single‐phase Sheppard–Taylor power factor correction (PFC) rectifier. Compared to conventional PFC rectifiers, this converter provides a better network current tracking and reduces voltage stresses. Furthermore, it proves an ability to control a detuning phenomenon from which most of the PFC circuits suffer. This yields a significant enhancement in the power quality (PQ). The proposed control scheme consists of two successive controllers: (1) a fuzzy logic controller (FLC) for the voltage loop, which ensures a good robustness to changes in parameters; and (2) a predictive current control (PCC), which allows to achieve a sinusoidal network current and improves the power factor. The proposed configuration is examined by simulation and implemented in real‐time using a dSPACE DS1104 controller board. The robustness of the proposed control is tested under different loading conditions and output voltage variations. The obtained results validate the superiority of the applied approach in terms of unity power factor (0.998), THD of the network current (3.7%, approximately), and fast dynamic response.

Model predictive sliding mode control of six‐phase induction motor using nine‐switch converter

“…The control goal error terms are usually involved in the cost function, and an exhaustive search is employed to decide the optimum control actuation. The control sequence that offers the minimum cost value is recorded as an optimum input 18–20 . Fundamentally, the traditional MPC evaluates only the penalization terms, the error between the reference signals and the predictions, and no closed‐loop stability investigation is performed 21–24 .…”

“…The control sequence that offers the minimum cost value is recorded as an optimum input. [18][19][20] Fundamentally, the traditional MPC evaluates only the penalization terms, the error between the reference signals and the predictions, and no closed-loop stability investigation is performed. [21][22][23][24] Due to the lack of stability assessment in the optimization procedure, the traditional MPC formulation may agonize over the system instability in case of deviation from the nominal operating points.…”