Monthly load forecasting is the basis for medium and long-term operation of power system and development of marketing work

and probabilistic power load forecasting can portray medium and long-term uncertainty

and better support the new type of power system load assessment and regulation strategy development. In this context

the medium and long-term probabilistic forecasting method is studied with the system load as the research object

and the medium and long-term probabilistic forecasting method based on fine-grained data fusion is proposed. Firstly

an hourly multiple linear regression model is established to model the fine-grained loads based on the influencing factors

and then the fine-grained forecasts under different scenarios are generated based on the different predicted values of the influencing factors. Secondly

according to the“bottom-up”temporal hierarchy coordination strategy

monthly aggregation is performed for each scenario

and monthly load forecasts are generated for different hierarchical regions to form probabilistic forecasts. Finally

the effectiveness of the method is verified by taking load data of a region in eastern China and its subordinate areas as an example.