Dynamic Voltage Restorer (DVR) is a custom power gadget utilized as a successful arrangement in shielding touchy burdens from voltage aggravations in power dissemination frameworks. The productivity of the control system, that directs the exchanging of the inverters, decides the DVR effectiveness. Corresponding Integral-Derivative (PID) control is the general method to do that. The power quality rebuilding capacities of this controller are constrained, and it produces critical measure of music – all of which comes from this straight procedure’s application for controlling non-direct DVR. As an answer, this paper proposes an Artificial Neural Network (ANN) based controller for improving rebuilding and sounds concealment abilities of DVR. A point by point examination of Neural Network controller with PID driven controller and Fuzzy rationale driven controller is additionally represented, where the proposed controller exhibited unrivaled execution with a unimportant 13.5% Total Harmonic Distortion.
Fig. 1 Simulation model for sag mitigation with ANN controller.
EXPECTED SIMULATION RESULTS:
Fig.2 Three phase sag mitigation based on ANN controlled DVR. (a) Instantaneous voltage at stable condition; (b) Instantantaneous voltage when sag occurs; (c) Voltage required to mitigate voltage sag; (d) Output voltage of the inverter circuit; (e) Generated PWM for inverter; (f) Instantaneous voltage after voltage restoration.
Fig 3. Restored Voltage Using (a) PID controller; (b) Fuzzy controller; (c) ANN controller; (d)THD comparison: the least THD can be seen at ANN based DVR, the range of the harmonics is also truncated by a huge amount by this method.
DVRs are a famous decision for upgrading power quality in power frameworks, with a variety of control framework on offer to drive these gadgets. In this paper, utilization of ANN to work DVR for giving preferable execution over existing frameworks to relieve voltage list, swell, and music has been illustrated. Issue articulation and hypothetical foundation, structure of the proposed strategy, preparing system of the ANN utilized have been portrayed in detail. Recreation results demonstrating the DVR execution amid voltage droop have been exhibited. Examination of the proposed technique with the well known PID controller, and nonlinear Fuzzy controller has been completed, where the proposed ANN controller showed up as the best choice to reestablish framework voltage while alleviating THD to the best degree.