A customized power device called a Dynamic Voltage Restorer (DVR) is an efficient way to safeguard delicate loads in power distribution networks against voltage fluctuations. The effectiveness of the control method used to switch the inverters is what defines the DVR's effectiveness. The general method for doing such is called Proportional-Integral-Derivative (PID) control. Because this linear approach is used to regulate a non-linear DVR, its ability to restore the power quality is constrained, and it generates a substantial quantity of harmonics. This study suggests an Artificial Neural Network (ANN) based controller as a remedy for improving the restoration and harmonics suppression capacities of DVR. The suggested controller showed improved performance with a mere 13.4% Total Harmonic Distortion (THD) in a comprehensive comparison of neural network controllers with PID-driven controllers and fuzzy logic-driven controllers.