A power load splitting method for dividing seasonal attributes

Abstract

The invention relates to a power load splitting method for dividing seasonal attributes. The method is characterized in that on the basis of seasonal division of meteorological data, a set of visual seasonal attribute judgment rule is determined to redivide a heating season, a transition season and a refrigeration season, and the influence of meteorological parameters on loads, including temperature, relative humidity, wind speed and air enthalpy values, is comprehensively considered. The method provided by the invention has the advantages and positive effects that the cage division of seasonal attributes by a traditional degree-of-day method is avoided, and a weighted seasonal attribute method is adopted, so that the division of heat supply, refrigeration and transition seasons is more reasonable and clear, and the reference of load splitting is more accurate; the load is divided into the basic load based on the average of the transition season and the meteorological sensitive load generated by heat supply or refrigeration, and then the basic load and the meteorological sensitive load are processed and predicted respectively, so that the relationship of different characteristics of the load is highlighted, the rule is easily found from the complex load curve containing components, and a more accurate prediction result is obtained.

Description

technical field [0001] The invention relates to an electric load splitting method for dividing seasonal attributes. The research object is the electric load of buildings. The electric load after splitting has obvious characteristics, which can improve the prediction accuracy of the building load. Background technique [0002] With the development of intelligent algorithms and machine learning, accurate prediction of urban power load based on effective algorithms is of great significance for guiding urban load power management and energy-saving strategies. Different user types have different power consumption characteristics. For example, the relatively intensive power load of office buildings occurs from 7:00 to 18:00 on weekdays, and the energy consumption of commercial buildings is mainly concentrated from 8:00 to 21:00 every day. The electricity consumption of the secondary industry presents different characteristics of electricity consumption according to different proce...

AI Technical Summary

Problems solved by technology

[0003] Therefore, if the factors affecting the load cannot be fully considered and the load can be split reasonably, the accuracy and generalization of the forecast will be affected, so that a comparative Accurate Forecast Results

Images