ISSN 1000-0526
CN 11-2282/P
CN 11-2282/P
Archive > Volume 47 Issue 7 > 2021,47(7):872-879. DOI:10.7519/j.issn.1000-0526.2021.07.010 Prev Next
Study on Segmental Method and Prediction Effect of Daily Electricity Load Based on Temperature and Humidity
- Article
- Figures
- Metrics
- Preview PDF
- Reference
- Related
- Cited by
- Materials
Abstract:
Meteorological condition is one of important factors in influencing electricity consumption, and is widely used in electricity load forecasting model. This study uses daily electricity load and meteorological data in Tianjin during 2014-2018 and analyzes the relationship between electricity and meteorological elements. The results show that the relationship between daily electricity load and temperature and relative humidity is nonlinear correlation with “U” shape. With the increasing of relative humidity, the temperature threshold between the comfort zone and the cooling zone is obviously offset to the low temperature side. Thus, the relative humidity is joined into the electricity load relational model. Based on the slope of the non-linear fitting curve between electricity load and temperature and relative humidity, a new method which considered the temperature and humidity effect of electricity consumption segmentation is proposed. It is found that the new method can effectively promote the fitting level of electricity load. In liner model, compared with the “V” segmented method, the root mean square error (RMSE) and mean absolute percentage error (MAPE) values decrease by 1.562 GW·h and 0.546%, respectively. For transition area between comfort zone and cooling zone (21.1℃ to 26.2℃), the RMSE and MAPE values decrease by 0.759 GW·h and 0.215% compared with the traditional “U” segmented method, while the RMSE and MAPE values decrease by 0.647 GW·h and 0.209% in the nonlinear model. This shows the stable prediction effect of different models. Thus, this “U” segmented method based on temperature and humidity effect can effectively improve the accuracy of daily electricity load forecasting.
Keywords:
Project Supported:
Clc Number:
Mobile website
