Power and Voltage Estimation of Unobserved Bus on Distribution Network Using ANFIS Algorithm with the Modified PSO-GA Hybrid

Abstract

Power and voltage estimation is a procedure for estimating the voltage of all buses in an electric power system based on measurements made on several buses. The voltage and the current on the bus are measured by a measuring device called the Phasor Measurement Unit (PMU). The Surabaya-Indonesia Bendul-Merisi distributed network has become the object of research for optimizing PMU placement by dividing into 3 clusters and placing 3 PMU on selected buses as observed buses. In this study, power and voltage estimation is done on buses that are not installed PMU (unobserved bus) using ANFIS algorithm with modified PSO-GA hybrid. This algorithm is used because of its ability to get predictive values with high accuracy and speed in achieving convergence. Input data for the estimation process are the voltage and current of all the buses. There are 10 data sets, divided into 7 training data and 3 test data. Estimation results compared with measurement results show a high degree of accuracy. The average accuracy of estimation results for power to 3 clusters is 99.968%, while the estimated accuracy for voltage is 99.758%.