Method for forecasting load based on historical data mining of heat-supply network

Abstract

The invention provides a method for forecasting load based on historical data mining of a heat-supply network, belonging to the heating ventilating and air conditioning (HVAC) field and solving the problems that unreasonable data are not eliminated, revision is not performed based on real-time heat supply conditions and forecasting results are unreasonable and are poor in precision in the existing method for forecasting load. The method of the invention comprises the following steps: 1, acquiring a set of historical data points of N-numbered heat-supply working conditions; 2, fitting the outdoor temperature set acquired in step 1 and an actual heating load set into a primary curve by adopting a least square method; 3, acquiring the heating load set subjected to primary fit corresponding to the outdoor temperature set T0; 4, acquiring a set of absolute values of deviations; 5, eliminating 10% of data points in the set of the absolute values of deviations in accordance with an order from small to large; 6, determining whether executing the screening steps from step 1 to step 5 for three times, if not, going back to step 1 and if so, executing step 7; 7, fitting all data points in the screened data point sets by the least square method to obtain a primary curve; and 8 carrying out twice revision to acquire a load forecasting curve so as to forecast the load.

Description

technical field [0001] The invention relates to a load forecasting method based on historical data mining of a heat network, and belongs to the field of heating and ventilation. Background technique [0002] The existing load forecasting methods of the heating network system are basically borrowed from the load forecasting methods in the power system. These forecasting methods use the historical data of the heating network as reasonable data for system identification and then use the identified model for load forecasting. . However, the difference between the heating network system and the power system is that most of the current heating system heat users do not use heat on demand, but can only passively receive heat. Therefore, the heat supplied by the heat source to the heat user is not necessarily an appropriate amount, and there may be situations of excessive heat supply and insufficient heat supply. However, the existing load forecasting methods do not eliminate unrea...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to solve the problem that the existing load forecasting method does not eliminate unreasonable data, does not correct based on real-time heating conditions, and the forecast results are unreasonable and the accuracy is poor. load forecasting method

Images