The accurate forecasting of photovoltaic (PV) power generation is critical for smart grids and the renewable energy market. In this paper, we propose a novel short-term PV forecasting technique called the space-time convolutional neural network (STCNN) that exploits the location information of multiple PV sites and historical PV generation data. The proposed structure is simple but effective for multi-site PV forecasting. In doing this, we propose a greedy adjoining algorithm to preprocess PV data into a space-time matrix that captures spatio-temporal correlation, which is learned by a convolutional neural network. Extensive experiments with multi-site PV generation from three typical states in the US (California, New York, and Alabama) show that the proposed STCNN outperforms the conventional methods by up to 33% and achieves fairly accurate PV forecasting, e.g., 4.6–5.3% of the mean absolute percentage error for a 6 h forecasting horizon. We also investigate the effect of PV sites aggregation for virtual power plants where errors from some sites can be compensated by other sites. The proposed STCNN shows substantial error reduction by up to 40% when multiple PV sites are aggregated.

No abstract

No abstract

In an era of high penetration of renewable energy, accurate photovoltaic (PV) power forecasting is crucial for balancing and scheduling power systems. However, PV power output has uncertainty since it depends on stochastic weather conditions. In this paper, we propose a novel short-term PV forecasting technique using Delaunay triangulation, of which the vertices are three weather stations that enclose a target PV site. By leveraging a Transformer encoder and gated recurrent unit (GRU), the proposed TransGRU model is robust against weather forecast error as it learns feature representation from weather data. We construct a framework based on Delaunay triangulation and TransGRU and verify that the proposed framework shows a 7–15% improvement compared to other state-of-the-art methods in terms of the normalized mean absolute error. Moreover, we investigate the effect of PV aggregation for virtual power plants where errors can be compensated across PV sites. Our framework demonstrates 41–60% improvement when PV sites are aggregated and achieves as low as 3–4% of forecasting error on average.

To save energy from cellular networks or to increase user-perceived performance, studying base station (BS) switching on-off is actively ongoing. However, many studies focus on the tradeoff between energy efficiency and user-perceived performance. In this paper, we propose a simple technique called cell flashing. Cell flashing means that base stations are turned on and off periodically and rapidly so that, when one base station is turned on, the adjacent base stations which make interferences are always off. Thus, both energy efficiency and cell edge user performances can be improved. In general, switching off base stations to save energy can lead to longer file download time (or delay) to customers. Using flow-level dynamics, we analyze average delay and energy consumption of cellular networks when cell flashing is used. We show that both of total energy consumption and average flow-level delay decrease in the case of small cells. Extensive simulations confirm that cell flashing can significantly save the energy of the base stations, e.g., by up to 25% and, at the same time, reduce the average delay by up to 75%.

Machine learning-based time-series forecasting has recently been intensively studied. Deep learning (DL), specifically deep neural networks (DNN) and long short-term memory (LSTM), are the popular approaches for this purpose. However, these methods have several problems. First, DNN needs a lot of data to avoid over-fitting. Without sufficient data, the model cannot be generalized so it may not be good for unseen data. Second, impaired data affect forecasting accuracy. In general, one trains a model assuming that normal data enters the input. However, when anomalous data enters the input, the forecasting accuracy of the model may decrease substantially, which emphasizes the importance of data integrity. This paper focuses on these two problems. In time-series forecasting, especially for photovoltaic (PV) forecasting, data from solar power plants are not sufficient. As solar panels are newly installed, a sufficiently long period of data cannot be obtained. We also find that many solar power plants may contain a substantial amount of anomalous data, e.g., 30%. In this regard, we propose a data preprocessing technique leveraging convolutional autoencoder and principal component analysis (PCA) to use insufficient data with a high rate of anomaly. We compare the performance of the PV forecasting model after applying the proposed anomaly detection in constructing a virtual power plant (VPP). Extensive experiments with 2517 PV sites in the Republic of Korea, which are used for VPP construction, confirm that the proposed technique can filter out anomaly PV sites with very high accuracy, e.g., 99%, which in turn contributes to reducing the forecasting error by 23%.

No abstract

Collaborative filtering (CF) is a widely used technique in recommender systems by automatically predicting the user's latent interests based on many users' historical rating data. To improve the performance of the CF-based recommender systems, users' rating data should be pre-processed to avoid noise and enhance data reliability. Many researchers studied anomaly detection to remove malicious noise caused by shilling attacks, but anomalies can still exist in non-attacked real user data, which is called natural noise, as the ratings of users can be impacted by unpredictable factors such as other users' ratings and anchoring bias. In this paper, we propose an autoencoder-based recommendation system for exploiting the ability of both anomaly detection and CF. The proposed system detects the natural noise in the rating data based on the reconstruction errors after training. By removing the detected natural noise, CF can predict the unrated ratings with noise-free data. Our experiments show that the proposed model showed better performance than the traditional method by reducing the error by up to 5% compared to the method that does not consider natural noise detection and reducing the error by up to 4% compared to the conventional rating classification based natural noise detection methods.