Study On Power Load Forecasting Based On Tieling

With the deepening of Chinese reform and opening up and growing economicrestructuring, electricity market is constantly changing, electrical energy supply anddemand is increasingly notable, power load forecasting has become increasinglyimportant and requirement for the prediction accuracy increases. Medium-long termpower load forecasting, as a planning basis for the grid, has the greater need for intensiveresearch. Medium-long term power load forecasting is important to ensure a reliablesupply of electricity and the safety and economic operation of the system, its accuracywill have an impact to invest in the grid, the topology of the network and even rationalityof run.In the actual network planning, there are a lot of load forecasting methods in use,such as the traditional time-series forecasting methods, linear regression method. Buttraditional forecasting methods have some shortcomings. They lead to its accuracy isdifficult to meet the requirements of modern power systems. Therefore, some modernforecasting methods have emerged, gradually also be applied to the actual load forecast.This paper uses several methods, mainly uses combination forecasting method, inforecasting TieLing power load in2013-2018. The most effective means of improvingthe accuracy of load forecasting is combination forecasting method. With TieLing actualbackground, conduct medium-long term power load forecasting, first introduced severalsingle load forecasting methods, then according to the load characteristics of TieLingarea, choose a more accurate load forecasting methods, single prediction methods werechosen preferred combination. The fitting model of single forecasting methods andcombinations of methods were established, to predict the load of Tieling from2013to2018, analysis the forecasting results, then, come to the conclusion that the combinationmethod is better than the single method. This method is more scientific and morepractical.