GIS-assisted modeling of wind farm site selection based on support vector regression

The optimal siting of wind farms is a multi-attribute problem, and different objectives and constraints like economic, environmental, social, and technical should be taken into account. The widely-used category of site selection methods, multi-criteria decision making (MCDM), suffers from local scoring despite its numerous ad-vantages. In this paper, support vector regression (SVR) is implemented to add the global scoring capability to MCDM by using a global-scale span of decision criteria. Here, the weighting of the decision criteria is obtained by the analytic hierarchy process (AHP), and then an SVR is utilized to appoint a score to any candidate site location. The proposed procedure has been used to study potential wind sites in Iran. In this regard, the study was conducted through three scenarios based on the natural and artificial potential of candidate sites. This study reveals that there are several vast areas, especially in eastern Iran, having high wind power density and flat terrain that are ready to be utilized as an endless source of clean energy provided that proper investment is made in power transmission lines and road networks.

Automated summary

In this paper, the authors propose a procedure for wind farm site selection using support vector regression (SVR) to incorporate global scoring capability to multi-criteria decision making (MCDM). The weighting of decision criteria is done using analytic hierarchy process (AHP), and an SVR model is used to assign scores to candidate site locations. The study conducted in Iran reveals the potential of vast areas, particularly in eastern Iran, with high wind power density and flat terrains that can be utilized as a sustainable source of clean energy with proper infrastructure development.