Low-voltage power demand forecasting using K-nearest neighbors approach

摘要: Demand response in the low-voltage domain has been ofter proposed to mitigate volatility of renewable energy supply. Therefore, an accurate demand forecast this is indispensable effectively manage balancing power. At same time, load profile based forecasting techniques, such as standardized profiles commonly used distribution grid, are inadequate for purpose. In article, we introduce a novel short-term model on K-nearest neighbors approach. Using historic smart meter data only input, it forecasts next day without any explicit knowledge consumer. our requires no manual setup while being parametrized automatically. Its accuracy shown be superior individual technique various samples low voltage end-consumers, and their aggregation group size. This makes viable wide-area application domain.