Research On Heating Load Forecast Based On Time Genetic Properties

In recent years,with the development of electronic computer technology,centralized heating system has been moving forward in the direction of intelligence,informatization and automation,the heating system is gradually changed from the coarse state regulation to the refined adjustment and the heating load is the basis of refinement adjustment.Using simple and feasible theoretical method with the existing historical data to accurately predict the heat consumption is the key to to realize refined adjustment of intelligent heating system.Based on the theory of architectural thermal characteristics and previous research results,this paper presents a prediction theory and method based on time genetic characteristics in which the building historical heat load and outdoor air temperature are independent variables.Firstly,the concept of the "time genetic period of building heat load" is proposed.And then,due to different thermal properties have different time genetic periods of building heat load,we use DeST software to analyze time genetic period of building heat load under different thermal characteristics,different heating periods,different heating temperatures and different geographical location;Finally,using MATLAB software to establish the model of outdoor temperature and historical heat load as the independent variables,and the model is used to predict the thermal load of different heating periods and different areas.According to the law of relative error fluctuation of heat load prediction value and measured value,the number of days of historical thermal load in the model is divided into three stages,which are defined as related fluctuation period,related stationary period and related weakening period,and the prediction accuracy of related stationary period is highest.The results verified the accuracy of the genetic characteristics of thermal load time and the time genetic period of the DeST software simulation.Finally,by comparing the accuracy of the heat load forecast results for the first day,third day,and the seventh day,it is concluded that the prediction model for the first day heat load forecast accuracy within 5%,which proved that the prediction model of the reliability and provide theoretical guidance for practical heating system heat load forecasting.