全球变化知识资源中心

The use of renewable energy resources, especially wind power, is receiving strong attention from governments and private institutions, since it is considered one of the best and most competitive alternative energy sources in the current energy transition that many countries around the world are adopting. Wind power also plays an important role by reducing greenhouse gas emissions and thus attenuating global warming. Another contribution of wind power generation is that it allows countries to diversify their energy mix, which is especially important in countries where hydropower is a large component. The expansion of wind power generation requires a robust understanding of its variability and thus how to reduce uncertainties associated with wind power output. Technical approaches such as simulation and forecasting provide better information to support the decision-making process. This paper provides an overview of how the analysis of wind speed/energy has evolved over the last 30 years for decision-making processes. For this, we employed an innovative and reproducible literature review approach called Systematic Literature Network Analysis (SLNA). The SLNA was performed considering 145 selected articles from peer-reviewed journals and through them it was possible to identify the most representative approaches and future trends. Through this analysis, we identified that in the past 10 years, studies have focused on the use of Measure-Correlate-Predict (MCP) models, first using linear models and then improving them by applying density or kernel functions, as well as studies with alternative techniques, like neural networks or other hybrid models. An important finding is that most of the methods aim to assess wind power generation potential of target sites, and, in recent years the most used approaches are MCP and artificial neural network methods. (C) 2019 Elsevier Ltd. All rights reserved.