‘Design and optimization of axial flux generators using evolutionary algorithms for grid-tied small wind turbines in rural electrification applications’


Subject of the present work is the study and optimization of small axial flux coreless permanent magnet generators for wind energy rural applications, at a range of nominal power from 3 to 6 KW.

An introduction is made on the principles of appropriate technology applications while the basic concepts of energy transformations of the wind generator are presented. A comparison between AFPM and RFPM is conducted and a theoretical analysis is also presented.

The complete sizing process of these generators is presented, including the determination of the nominal operating conditions and the total set of dimensions of the generator’s parts. The dimensions of coils and magnets are determined as well as the total cost, weight and volume of the generators which are going to be use in the process of optimization.

According to the previous approach the design is optimized with the Particle Swarm Optimization (PSO) and Homogeneous Particle Swarm Optimization (HPSO) methods while analysing the theory of these optimizing algorithms. The pole arc to pole pitch ratio and the ratio of the internal to the external diameter are optimized for lower cost, mass, volume and higher efficiency. Also combinations of these criteria are optimized with the HPSO method. The results of these combinations and a comparison for magnet grades N40 and N45 are presented.

Further on, the thermal analysis of a 3 KW is examined. The relevant theory is analysed and the thermal model of lumped parameter is used. The results are compared with the experimental thermal measurements.

Finally, conclusions are drawn and proposals for improvements and further work are presented.

Keywords: appropriate technology, low cost, small wind turbines, axial flux generators, permanent magnets, cost optimization, mass optimization, Particle Swarm Optimization (PSO), Homogeneous Particle Swarm Optimization (HPSO), objective function, magnet grade, thermal analysis, lumped parameter thermal model.

Author: Achilleas Tsitsimelis

Responsible PhD: Kostas Latoufis / latoufis@power.ece.ntua.gr

Supervising Professor: Nikos Hatziargyriou / nh@power.ece.ntua.gr

PDF: Full version (Greek) and Short version (English)

This entry was posted in Uncategorized. Bookmark the permalink.