Progressively-adjusted load prediction method for directly-heated electric boiler in residential area in northeast China

Abstract

The invention relates to a progressively-adjusted load prediction method for a directly-heated electric boiler in a residential area in northeast China, and provides a new load prediction method for apower system, so that historical experience can be summarized, and progressive adjustment can be carried out according to a current data value, and accurate load prediction is realized. Through the progressively-adjusted load prediction method, the load of the directly-heated electric boiler in the residential area in the northeast China can be accurately predicted, and the progressively-adjustedload prediction method plays an important role in power grid safety management, wind power peak regulation and power grid scientific management.

Description

technical field [0001] The invention relates to a progressively adjusted load forecasting method for direct-heating electric boilers in residential areas in Northeast China, provides a new load forecasting method for electric power systems, and belongs to the technical field of electric boiler load management. Background technique [0002] The direct heating electric boiler is a widely used heating technology in residential areas in Northeast my country, and heating accounts for a large proportion of electricity consumption in winter; therefore, accurate prediction of the direct heating electric boiler load can be used to plan the power grid in advance. It plays a very important role in power grid security management, wind power peak regulation, and power grid scientific management. [0003] To predict the load of direct-heating electric boilers, the technical means commonly used in the industry include: 1. Directly use the load situation in the same period of previous years ...

AI Technical Summary

Problems solved by technology

[0003] To predict the load of direct-heating electric boilers, the technical means commonly used in the industry include: 1. Directly use the load situation in the same period of previous years as the current forecast result. Such methods can only "guess" the approximate magnitude of the load However, it is impossible to predict the accurate value, and it is difficult to serve as the basis for scientific management

Second, use neural network, decision tree and other models to learn historical data to obtain a regression model from outdoor temperature to load, and then use the regression model to predict; such methods can predict the specific value of load, however, the electricity consumption in residential areas is still Affected by various uncertain factors such as pipeline aging, increase in occupancy, and aging of insulation materials, in addition, the water outlet and water inlet temperature settings of the boiler may also change with policies and electricity prices, so the regression model will change over time. It is becoming more and more inaccurate, especially in newly built residential areas, the simple regression analysis model is difficult to achieve better results

Images