IMPROVING POWER QUALITY UNDER SECTION 73 OF THE ELECTRICITY ACT, 2003 THROUGH DYNAMIC VOLTAGE RESTORER WITH PROPOSED APPROACH

Abstract

A number of effects that alter electrical waveforms are arising as a result of the growth of nonlinear loads and electronic gadgets in the distribution network. In order to get over these obstacles, power demand must be satisfied in addition to improving distribution system power quality (PQ) performance. This manuscript presents a hybrid approach for voltage control of a dynamic voltage restorer (DVR). The hybrid technique is the combination of Dung Botox Optimization Algorithm (BOA) and Dual Attention Graph Convolutional Network (DAGCN) and then it is together named as BOA-DAGCN technique. The primary objective is to improve power quality along with DVR in order to achieve a global minimum error and fast dynamic response of the proposed controller.  The proposed BOA method is used to optimize the gain parameters of the PI controller. The DAGCN is used to predict the load demand. Finally, the performance of the proposed technique is implemented in MATLAB working platform and compared to various existing approaches. Furthermore, the existing techniques such as Cuckoo search (CSA), Flower pollination Algorithm (FBA), and Grey wolf optimizer (GWO).  From the result, the error of proposed method is 1.0936, 1.0158 and 0.1505. The existing CSA is 1.3373, 1.2559 and 0.1324, FBA is 1.4845, 1.3863 and 0.1483 and GWO is 1.5784, 1.4892 and 0.5883.