Enhanced multi-objective optimization algorithm for renewable energy sources: optimal spatial development of wind farms

Abstract

A new approach for treating multi-objective spatial optimization problems is introduced in this study, aiming at deriving the optimal spatial allocation of Wind Farms on a Greek Island (Lesvos). This work builds on the knowledge gained from numerous applications of multi-objective genetic algorithms, either for spatial planning purposes or for other engineering-related topics, by incorporating modified genetic operators and sophisticated planning criteria. Hence, a stand-alone genetic optimizer was developed that incorporates the controlled non-dominated sorting genetic algorithm-II (CNSGA-II), in which the user can model all planning criteria and constraints for every spatial entity to be allocated, and handle the genetic solver via a built-in computational framework that permits the analysis of large terrains. The presented paradigm provides interesting findings for the optimal development of renewable energy sources projects whose spatial allocation is governed by conflicting criteria and strict constraints.

Keywords:

• multi-criteria analysis
• genetic algorithms
• spatial analysis
• spatial planning
• decision support system
• wind farms

Acknowledgments

The authors would like to thank all GIS experts from Mr. Economou’s consulting office for providing the necessary spatial data for the development of Lesvos’ spatial model, and Mr. Dimitrios Stathakis³ for his valuable feedback and guidance.

Notes

1. 1. Crowding distance metric: half of the perimeter of the maximum hypercube that can be allowed around a solution without including any other solution from the same Pareto front. A comprehensive guide for this metric can be found in K. Deb’s textbook (Deb 2001b, p. 248–252).

2. 2. It is possible to obtain the same score for a specific design objective from two different maps. Hence, the evolving spatial representations must be compared cell by cell.

3. 3. Assistant Professor in Urban and Spatial Planning using GIS and Remote Sensing, Department of Planning and Regional Development, University of Thessaly, Volos, Greece.