Power load prediction method considering epidemic situation event

Abstract

The invention relates to a power load prediction method considering an epidemic situation event. The method comprises the following steps: (1) constructing characteristic engineering reflecting the development situation and weather of the epidemic situation event; (2) establishing a load prediction model; and (3) training the prediction model by using the original data, and solving by using a gradient descent method. The method can realize real-time adjustment and autonomous learning, and is high in adaptability and high in prediction precision.

Description

technical field [0001] The invention belongs to the field of big data and artificial intelligence, and in particular relates to a power load forecasting method considering epidemic events. Background technique [0002] Power load forecasting is a series of forecasting work on power load. From the perspective of forecasting objects, electric load forecasting includes forecasting of future demand for electricity (power), forecasting of future electricity consumption (energy) and forecasting of load curves. Its main work is to predict the time distribution and spatial distribution of future power load, and provide reliable decision-making basis for power system planning and operation. [0003] Power load forecasting is an important daily work. Accurate load forecasting can improve the safety, reliability and economy of transmission and distribution network operation. Traditional power load forecasting is mainly based on the load data of the same period, and short-term and per...

AI Technical Summary

Problems solved by technology

When conducting power load forecasting, if only a simple relationship is used to reflect the relationship between power load and its influencing factors (such as weather, environment, economy, etc.), the forecast results will deviate far from the actual ones. If a model is established, and Because the respective models themselves contain uncertain factors, there are errors of different sizes

[0010] With the rapid development of computer technology, it is possible to adopt a large number of complex forecasting methods that are difficult to achieve with manual methods, and the means of load forecasting have gradually developed into the use of software forecasting, which greatly increases the methods and means of load forecasting. But there are still big problems in making accurate predictions

Images