Development of a cooling load prediction model for air-conditioning system operation control
Building cooling load prediction is of critical importance for achieving energy saving of air-conditioning system in high-rise buildings.It not only benefits the energy-efficiency operation of the air-conditioning system,but also great important for the system stability.Many techniques have been developed for building cooling load prediction and these methods are normally arranged into three categories: regress analysis,energy simulation and artificial intelligence.Different from the last two categories,which may have better prediction accuracy but be much complicated,the regression analysis methods are much practical for real applications.It is suggested to use the regression models in the air-conditioning system control for achieving a better operation.However,traditional regression models are often helpless to manage multi-parameter dynamic changes,and the effect of outliers in cooling load prediction has not been considered seriously,accuracy of building hourly cooling load prediction is not satisfied.To promote the application feasibility of regression cooling load prediction models,this study developed an efficient hourly cooling load prediction model based on sensitivity analysis and the traditional auto-regressive with exogenous(ARX)model.The developed cooling load prediction model kept the constitution of ARX model,but used different variables from sensitivity analysis as inputs,also the quadratic term of key variables was included in the model.The prediction accuracy of the developed was validated by compared it to the model established in EnergyPlus,which is a widely acknowledged software for energy modeling.Case studies showed that the deviations between the developed method and EnergyPlus were less than ±10%.The high accuracy ensured that the developed method could provide a reliable prediction for air-conditioning system control.Since the developed method is much less complicated than the energy simulation,it will be much suitable for real operational controls.
Cooling load prediction Operational control ARX model Regression analysis Sensitivity analysis
Chengliang FAN Yundan LIAO Yunfei DING Zhenbing CAI
School of Civil Engineering,GuangZhou University,Guangzhou 510006,China School of Civil Engineering,GuangZhou University,Guangzhou 510006,China;Guangdong Provincial Key Lab
国际会议
武汉
英文
80-85
2018-08-21（万方平台首次上网日期，不代表论文的发表时间）