Combined support vector machine based method for predicting power load of large city in medium and long term

Abstract

The invention relates to a combined support vector machine based method for predicting the power load of a large city in a medium and long term. The method is characterized by comprising following steps: step one, taking the subarea characteristic values and subarea area of medium and long term load of a large city as the inputs, and providing a load density based method for predicting the power load of the large city in a medium and long term; step two, taking historical annual load data as the input, and providing a trend extrapolation based method for predicting the power load of the largecity in a medium and long term; step three, taking expert experience and historical annular load data as the inputs, and providing an expert experience based average annual growth rate method for predicting the power load of the large city in a medium and long term; and step four, based on the results of the step one, two, and three, fitting the data by a support vector machine method, and providing a support vector machine based method for predicting the power load of the large city in a medium and long term. Three load prediction methods are fitted for a second time to obtain a predicted load value of a power network of a large city in a medium and long term.

Description

technical field [0001] The invention belongs to the technical field of load forecasting, and relates to a medium and long-term power load forecasting method, in particular to a medium and long-term power load combination forecasting method based on a support vector machine in large cities. Background technique [0002] The medium and long-term demand forecast of the power market is an important basis for power grid planning, investment and construction. With the implementation of my country's new power reform policy, emerging loads such as distributed power sources, microgrids, and electric vehicles will develop in a blowout manner. Habits have also undergone major changes with economic development. In the past, the medium and long-term forecasting method of the electricity market based on economic and social development factors and historical power consumption data as the main factors to forecast can no longer adapt to the current form of development, and it is difficult to p...

AI Technical Summary

Problems solved by technology

Some scholars combine econometrics and system dynamics to form a combined forecasting model, which integrates development factors, but does not consider the error of econometrics itself; some scholars also propose a variable weight gray forecasting model, which is simple to operate and relatively accurate. High, but the variable weight method is relatively backward and needs to be improved

[0004] Moreover, the existing research on urban mid- and long-term load forecasting methods is mainly based on single methods such as heuristic algorithms and intelligent algorithms, and rarely involves the mixed forecasting methods of multiple forecasting methods for medium and long-term loads in large-scale urban power grids.

Images