We propose a pooling-ensemble model for short-term residential load forecasting. Compared with traditional load forecasting, the frequent spikes and the volatile load patterns require to be considered in residential load forecasting. We propose a smoothing clustering method to obtain the daily load states of residents. Smoothing operation in clustering can abolish noise compared with directly clustering. We built a W3M to predict the states of the residents in next day by leveraging a day-skip Markov model and a week-skip Markov model. For the load forecasting task of each cluster, we select the most appropriate model from the basic models. The selected models are combined as an ensemble model where the weights are the probability of the predicted states. The proposed pooling-ensemble model can achieve better short-term load forecasting on two public datasets. Accurate load forecasting can be applied to home energy planning and demand response projects.

Short-term residential load forecasting is essential to demand side response. However, the frequent spikes in the load and the volatile daily load patterns make it difficult to accurately forecast the load. To deal with these problems, this article proposes a smoothing clustering method for daily load clustering and a pooling-ensemble model for one day ahead load forecasting. The whole short-term load forecasting framework in this article contains three steps. Specifically and first, the states of the residents are obtained by clustering the daily load curves with the proposed smoothing clustering method. Second, a weighted mixed Markov model is built to predict the probability distribution of the load state in the next day. Third, multiple predictors in the pooling-ensemble model are selected for different states and the load is forecasted by weighing the results of the multiple predictors based on the predicted states. Results of the case studies and comparison studies on two public ...