A Customized Binary-coded Genetic Algorithm for Wind-farm Layout Optimization

In this paper, a customized binary-coded genetic algorithm (GA) is proposed for wind-farm layout optimization problem. A flexible grid model that encodes the layouts to the binary-coded individuals of the population is optimized for the wind distribution. The bit-flipping GA operators are designed to handle the constraint of turbines number. The diversity in GA is measured by the variety of wind-farm layouts represented by the population. The novelty-search and duplicate individuals elimination are applied in the procedure of GA for diversity maintenance. The proposed method is compared to the methods with the traditional grid model and standard GA on the layout optimization problem for maximizing energy production. The simulations results demonstrate that the energy production of optimized layout is increased and the convergence of solution in repeated simulation is improved by the proposed algorithm.