Low-voltage ride-through operation of single-stage PV system via using recurrent fuzzy CMAC

Abstract

Advantages of the recurrent fuzzy cerebellar model articulation controller (RFCMAC) include local generalization capability, fast learning, simplicity of computation, and capability of solving dynamic problem. Normally, the tracking error and the derivative of the tracking error are fed into the controller to perform the control of command tracking. Firstly, in this study, a two-dimensional RFCMAC (2D-RFCMAC) is adopted to approach the command control of the active and reactive power for performing the low-voltage ride-through (LVRT) operation of a single-stage photovoltaic (PV) system. However, the 2D-RFCMAC requires larger memory due to the memory size increases exponentially with the number of inputs. Thus, a one-dimensional RFCMAC (1D-RFCMAC) with signed distance and input space repartition mechanisms is proposed to replace the 2D-RFCMAC to perform the LVRT operation of the single-stage PV system. The reduced input dimension and computation complexity make the 1D-RFCMAC more practical. Moreover, some experimental tests are presented to illustrate the effectiveness of the proposed controller. The results show that the control performance of the proposed 1D-RFCMAC is not only slightly better than 2D-RFCMAC but also has lower memory size and computation complexity.