Load Modeling For Residential Load Forecasting Studies

With the rapid development of social economy, people’s standard of living continuouslyimproved, the demand for electricity increased too. While public transformers is so quantity,wide geographical distribution, and load changes quickly along with strong randomness, it’svery difficult to predict the amount of each station area load growth, which brings a greatchallenge to public transformer operation, maintenance, and project planning.Currently, there is no comprehensive in-depth research on residential electricity demandtrend. Prediction on the development of the transformer load just stay at a preliminary, partial,short-term level, mostly experience to conjecture load growth trend. load forecastinginaccurate results in, on the one hand, light-load transformer capacity increase, caused a wasteof resources; on the other hand, even a large number of transformers are overloading oroverload, caused a great threat to the equipment.Therefore, it’s very important andsignificance to depth and scientific research the load variation of the transformer, and accurateforecast load growth.This paper presents a method for transformer load forecast. We take the grid as theresearch object to the prediction method, based on historical operating data. On initial wemake a load curve fitting, combined with GIS systems to extract spatial information, and thenstudy load density indicator method depth. The main factors which caused load changes arecombined, the purpose is to identify the variation of the load and predict the load on theparticular area in the future.At last, we select luotian, Yan Chuan, a total of nine grids which contain36publictransformers to analyze as an example, predict these grids load changes, and then,in the nextyear,verify the accuracy of the method with the measured data.