How many watts: A data driven approach to aggregated residential air-conditioning load forecasting

Due to the significant contribution of air-conditioning load towards total energy consumption in residential buildings, accurate modelling and forecasting of such load is key to effective demand-side energy management programmes. This paper suggests a data driven framework for 15 min-ahead AC load forecasting based on modern machine learning techniques that includes Support Vector Regression, Ensemble Trees, and Artificial Neural Network. To the end, it utilizes a correlation based feature selection method to identify information that is relevant for machine learning modelling. The effect of spatio-temporal features selection on prediction output and the effect of training data quantity on convergence characteristics were analysed and discussed. The effectiveness of the proposed approach is evaluated using a 20-household, half-year data set from an ongoing research testbed set up at the faculty housing units of Singapore University of Technology and Design. An linear combination method was proposed to combine models and the resulting model gave a mean absolute percentage error of 11.27%.

[1]  Hao Yu,et al.  Levenberg—Marquardt Training , 2011 .

[2]  Shira R. Horowitz,et al.  Forecasting residential air conditioning loads , 2014 .

[3]  J. Friedman Greedy function approximation: A gradient boosting machine. , 2001 .

[4]  Shuhui Li,et al.  Integrating Home Energy Simulation and Dynamic Electricity Price for Demand Response Study , 2014, IEEE Transactions on Smart Grid.

[5]  D. Basak,et al.  Support Vector Regression , 2008 .

[6]  Bernard Zenko,et al.  Is Combining Classifiers with Stacking Better than Selecting the Best One? , 2004, Machine Learning.

[7]  Raymond Chiong,et al.  Hybrid filter-wrapper feature selection for short-term load forecasting , 2015, Eng. Appl. Artif. Intell..

[8]  Hu Wei,et al.  Multi-variable time series forecasting for thermal load of air-conditioning system on SVR , 2015, 2015 34th Chinese Control Conference (CCC).

[9]  Ton J. Cleophas,et al.  Missing-data Imputation , 2022 .

[10]  Ping Zhang Model Selection Via Multifold Cross Validation , 1993 .

[11]  Ross Baldick,et al.  Demand response control of residential HVAC loads based on dynamic electricity prices and economic analysis , 2016 .

[12]  Li Xuemei,et al.  A Novel Air-Conditioning Load Prediction Based on ARIMA and BPNN Model , 2009, 2009 Asia-Pacific Conference on Information Processing.

[13]  Yik-Chung Wu,et al.  Load/Price Forecasting and Managing Demand Response for Smart Grids: Methodologies and Challenges , 2012, IEEE Signal Processing Magazine.

[14]  Vikas Chandan,et al.  Non-intrusive estimation and prediction of residential AC energy consumption , 2016, 2016 IEEE International Conference on Pervasive Computing and Communications (PerCom).

[15]  Yuan Lan,et al.  Ensemble of online sequential extreme learning machine , 2009, Neurocomputing.

[16]  Naci Kalkan,et al.  Solar thermal air conditioning technology reducing the footprint of solar thermal air conditioning , 2012 .

[17]  Wei Zhang,et al.  Aggregated Modeling and Control of Air Conditioning Loads for Demand Response , 2013, IEEE Transactions on Power Systems.