Automatic Identification of Repeating Energy Consumption Profiles
Наименование публикации | Automatic Identification of Repeating Energy Consumption Profiles |
Тип публикации | Статья |
Библиографическое описание | Automatic Identification of Repeating Energy Consumption Profiles |
Аннотация | Data preprocessing is the most essential and time consuming problem. The paper presents an original mathematical-statistical method for the analysis of the similarities and differences in electricity energy consumption time series in non-residential buildings. This method is based on the well-known chi-square Pearson statistics to test a number of hypothesis considering likeness and differences in energy consumption samples. The method is designed for identification of recurring consumption profiles and their calendar binding in automated mode. Identification of the profiles allows you to create important conclusions about segments of local modeling and forecasting. Thus, the analysis of similarities and differences precedes study of cause-effect relationships and allows to improve quality of energy consumption forecasting. The paper considers the formalization of the method necessary for its program implementation. The results of the identification of energy consumption profiles for shopping center in the Netherlands are presented. |
Ключевые cлова | energy consumption forecasting, time series analysis, energy consumption profile, statistical analysis, chi-squared test |
Год публикации | 2013 |
Автор(ы) |
Щербаков Максим Владимирович
Камаев Валерий Анатольевич
|
Электронная копия публикации | Загрузить |