Modeling an optimal wind turbine layout by application of evolutional algorithms
Date Issued
2019-10
Author(s)
Abstract
Optimization of wind farm layout concerning various parameters is a major point in planning
and will influence the revenue for the whole life of the installation. Besides the obvious impact of wind distribution also other parameters like connection costs and levelized costs of energy influence the optimum layout and have to be included in a realistic optimization algorithm.
In this disertation the sophisticated optimization of wind farm layout with of two fundamentally
different heuristic algorithms is investigated. To do so, detailed real-world data from an existing wind farm in Bogdanci, Macedonia is utilized by employing real wind farm data we are able to calibrate
model adequacy and ascertain a model that will serve as a referent guidance in the planning of future onshore wind farms. The major unique feature of the research is the simultaneous optimization taking into account all major technical influence and cost factors, including: (i) detailed and advanced models for power modeling due to bivariate distribution of wind speed and direction; (ii) accurate estimation of levelized cost of energy (LCOE); (iii) analysis of the shortest electrical interconnections among wind turbines and (iv) correction of hub height on each wind turbine in the wind farm with taking also the wake effect into consideration. Different layouts were designed using sophisticated algorithms for handling the resulting high-dimensional, highly non-linear optimization problem. In particular, a non dominated sorting genetic algorithm (NSGA) and a mixed-discrete particle swarm optimization
algorithm (MDPSO) were applied. Both optimization algorithms established bi-objective fitness
functions, in particular- minimizing the levelized cost of energy and maximizing the capacity factor. By comparing the results obtained with the existing layout, it is established that both optimization
algorithms are adequate in determination of wind power plant layouts. Results show also a remarkable improvement of 2.05% and 5.59% for levelized costs and capacity factor, respectively, compared to the as built wind farm layout.
and will influence the revenue for the whole life of the installation. Besides the obvious impact of wind distribution also other parameters like connection costs and levelized costs of energy influence the optimum layout and have to be included in a realistic optimization algorithm.
In this disertation the sophisticated optimization of wind farm layout with of two fundamentally
different heuristic algorithms is investigated. To do so, detailed real-world data from an existing wind farm in Bogdanci, Macedonia is utilized by employing real wind farm data we are able to calibrate
model adequacy and ascertain a model that will serve as a referent guidance in the planning of future onshore wind farms. The major unique feature of the research is the simultaneous optimization taking into account all major technical influence and cost factors, including: (i) detailed and advanced models for power modeling due to bivariate distribution of wind speed and direction; (ii) accurate estimation of levelized cost of energy (LCOE); (iii) analysis of the shortest electrical interconnections among wind turbines and (iv) correction of hub height on each wind turbine in the wind farm with taking also the wake effect into consideration. Different layouts were designed using sophisticated algorithms for handling the resulting high-dimensional, highly non-linear optimization problem. In particular, a non dominated sorting genetic algorithm (NSGA) and a mixed-discrete particle swarm optimization
algorithm (MDPSO) were applied. Both optimization algorithms established bi-objective fitness
functions, in particular- minimizing the levelized cost of energy and maximizing the capacity factor. By comparing the results obtained with the existing layout, it is established that both optimization
algorithms are adequate in determination of wind power plant layouts. Results show also a remarkable improvement of 2.05% and 5.59% for levelized costs and capacity factor, respectively, compared to the as built wind farm layout.
Subjects