Intelligent low-carbon park load prediction method and device based on actual meteorological conditions

The invention provides an intelligent low-carbon park load prediction method based on actual meteorological conditions, and the method comprises the steps: building a meteorological database, and generating a load data parameter set according to historical load data and historical meteorological data in the meteorological database; based on the load data parameter set and the historical meteorological data, determining a coefficient of a pre-established initial hourly multi-element nonlinear regression model of each building; determining a building type according to a data type selected by a user, and determining a hourly nonlinear function model of the to-be-tested park according to the coefficient of the initial hourly multi-element nonlinear regression model of each building and the building type; and inputting meteorological data of a region where the park to be measured is located into the hourly nonlinear function model to obtain a basic load, and correcting the basic load according to the environment c