Global Optimization of Wind Farms Using Evolutive Algorithms

Abstract

The design of a facility for wind power generation is a complex and multidisciplinary problem. The complexity of the problem derives mainly from the many interrelated variables and constraints or restrictions involved. Thus, the solution is usually obtained by heuristics after several cycles of trial and error, and it is heavily based on previous experience of the team planner. Evolutionary algorithms are efficient optimization techniques to tackle the problem of global optimization for wind farms by considering the turbines layout and electrical and civil infrastructure as a whole. The algorithm should evaluate each potential solution based on their economic returns over the entire period production of the wind farm, providing economic and financial information useful for prospective developers. Therefore, the algorithm needs to be driven by a thorough cost wind farm model that considers both the initial costs of acquisition and installation of equipment (initial investment) and the yearly cash flow. This cash flow is calculated as the difference between the incomes due to the energy selling and the ordinary maintenance and operation costs, along the whole lifespan of the wind farm. A final cost for the installation decommissioning and a residual value, after the facility production period, should also be considered. The content of this chapter is organized into five main sections. After an initial introductory section, the problem of wind farm design and planning is formulated. Then there is a brief section on the basics of evolutionary algorithms. Having discussed the problem and the optimization technique, the next section is devoted to integral wind farm optimization through evolutionary algorithms. This section includes a comparison with published works and a collection of new cases to test this new tool performance. The chapter ends with conclusions and references.