New Method of Modeling Daily Energy Consumption

At present, papers concerning energy consumption and forecasting are predominantly dedicated to various known techniques their combinations. On the other hand, research on load modeling methodologies is quite limited. This paper presents a new approach hourly using multivariate linear regression model. The proposed technique provides way accurately model day-to-day just few selected variables. number of data points required describe whole day’s depends demanded precision, which up user. self-configurable very fast. applied shows that four hours sufficient during remainder given day. We show for about 84% points, relative error below 2.5%, all does not exceed 7.5%. obtained mean uncertainty 1.72% in learning set, 1.69% 1.82% two testing sets, respectively. In addition, we conclude can also detect days with unusual consumption.