Short-term load forecasting based on human body amenity indicator and arithmetic of random forests

Abstract

In recent years, fog and haze are now becoming common in China, which produces unfavorable influence on the production and life of the residents, and further influences the electricity load and its trend. The severe air condition presents a challenge to power system short-term load forecasting. In this paper, air quality is brought into the comfortable degree index, which is used to power system short-term load forecasting during fog and haze occurrence. Basing on existing human body amenity indicator, a new concept of human body amenity considering AQI, temperature, humidity and wind is presented while analytic hierarchy process (AHP) is used to construct the human comfort index model. In short-term load forecasting, the human comfort index instead of various meteorological factors is taken as input, thus improving the precision of the power load forecast. The similar load days needed in power load forecasting are extracted by using the gray correlation analysis method, and based on that, the arithmetic of random forests is adopted in building forecasting model. The effectiveness and validity of proposed model and algorithm are verified by a city's practical load data and weather data of winter in North China area.