Short-term Electricity Forecasting Using Linear Regression
Novi Gusti Pahiyanti (a), Sigit Sukmajati (b)
(a) Departement Electrical Engineering, PLN Technical College
Jalan Lingkar Luar Barat Duri Kosambi, Cengkareng, Jakarta barat 11750
novi.gusti[at]sttpln.ac.id
(b) Departement Electrical Engineering, PLN Technical College
Jalan Lingkar LUar Barat Duri Kosambi, Cengkareng, Jakarta Barat 11730
Abstract
Short term electricity load forecasting is a very important factor in the planning and operation of an electric power system. The purpose of electricity load forecasting is that electricity demand and supply can be balanced. The pattern of consumption of different electrical loads at certain intervals at the Harapan Indah Gas Insulated Switchgear (GIS) substation. Making the problem of variations in the electricity load is not homogeneous. Lots of methods are used to produce accurate and precise electrical load forecasting. With the hope that the current load will be distributed precisely and according to the needs of consumers. In this study discusses the short-term electricity load forecasting by the Linear Regression method and the Energy Coefficient where, the two methods are compared to find the value of the percentage of small errors and very good accuracy. Calculations in this study use historical data on electricity loads at PT PLN (Persero) Jakarta Raya Main Distribution Unit in March 2019 on GIS Harapan Indah. The results showed that the average error ratio from Monday to Sunday was 7.92% with an accuracy value of 92.08% for Linear Regression and 10.46% with an accuracy value of 89.54% for the Energy Coefficient
Keywords: Short-term forecasting, linear regression, energy coefficient percentage error
Topic: Electrical Engineering