Saferay Pte. Ltd.

Singapore

Saferay Pte. Ltd.

Singapore
SEARCH FILTERS
Time filter
Source Type

Yang D.,Agency for Science, Technology and Research Singapore | Quan H.,Institute of Chemical and Engineering Sciences, Singapore | Disfani V.R.,University of California at San Diego | Liu L.,Saferay Pte. Ltd.
Solar Energy | Year: 2017

To integrate solar energy into the electricity grid reliably and efficiently, different forecasting techniques are used to produce forecasts at various forecast horizons using data of different levels of aggregation. For example, power generated by distributed photovoltaics (PV) can be disaggregated in a geographical hierarchy into transmission zones, distribution nodes, PV plants, subsystems and inverters; forecasts are required at all of these levels to facilitate different power system operations and power plant management. Due to the different information sets visible to different players in the hierarchy, such as PV plant owners and independent system operators, the forecasts produced at different aggregation levels are often not optimal. Furthermore, these forecasts are aggregate inconsistent, i.e., forecasts made using data collected at lower levels do not add up exactly to the forecasts made using higher level data. In this paper, the state-of-the-art forecast reconciliation techniques are explored. By minimizing the trace of forecast error covariance matrix, the base forecasts obtained across the hierarchy can be optimally reconciled. The reconciled forecasts not only aggregate consistently, and thus lead to collaborative decision making, but also improve the base forecasts at each level significantly. The merit of the reconciliation framework goes to the fact that it does not require any additional information other than those base forecasts. Furthermore, forecast improvements are independent of the base forecasting methods. In other words, once the base forecasts are generated with our favorite methods, the reconciliation will most likely improve those forecasts further. The reconciliation techniques can be applied to a variety of hierarchies with different forecast horizons. The empirical part of this work considers two examples, namely, an NWP-based day-ahead forecast reconciliation over 318 simulated PV plants in the state of California, and a spatio-temporal statistical 1-min-ahead forecast reconciliation over 17 irradiance sensors on the Oahu Island, Hawaii. It is shown that reconciliation benefits both the gird operators and PV system owners by bringing more accurate forecasts, and thus motivates information sharing in an electricity grid. © 2017 Elsevier Ltd


Yang D.,Singapore Institute of Manufacturing Technology | Dong Z.,Solar Energy Research Institute of Singapore | Lim L.H.I.,University of Glasgow | Liu L.,Saferay Pte. Ltd.
Solar Energy | Year: 2017

In solar engineering, we encounter big time series data such as the satellite-derived irradiance data and string-level measurements from a utility-scale photovoltaic (PV) system. While storing and hosting big data are certainly possible using today's data storage technology, it is challenging to effectively and efficiently visualize and analyze the data. We consider a data analytics algorithm to mitigate some of these challenges in this work. The algorithm computes a set of generic and/or application-specific features to characterize the time series, and subsequently uses principal component analysis to project these features onto a two-dimensional space. As each time series can be represented by features, it can be treated as a single data point in the feature space, allowing many operations to become more amenable. Three applications are discussed within the overall framework, namely (1) the PV system type identification, (2) monitoring network design, and (3) anomalous string detection. The proposed framework can be easily translated to many other solar engineer applications. © 2017 Elsevier Ltd

Loading Saferay Pte. Ltd. collaborators
Loading Saferay Pte. Ltd. collaborators