Improved Core Geometry of a Low Power Modular Wind Generator Using the Real Coded Genetic Optimization Technique

Renewable energy sources such as wind, solar, geothermal and tides have increasingly been used to generate electricity. Among these energy sources wind power has gained enormous attention in many parts of the world where no fossil fuels exist and energy requirements are highly demanded. The efficiency of wind power mainly depends on wind produced and a generator converting wind power to electricity. The efficiency of a typical wind power generator are usually related with its magnetic performance based on magnetic core geometry. Furthermore, this geometry usually has an effect on flux density distribution hence in some cases the flux overlaps in regions near the corners of the stator and the rotor cores; thereby leading to an increase of magnetic losses and decrease of magnetic performance. In this paper, conventional magnetic core geometry of a directly driven and modular permanent magnet low power generator is reshaped using 2-D magnetic field analysis and a genetic algorithm method. This has allowed us to obtain improved core geometry on flux distribution especially the magnetic performance of the generator from various aspects. The simulation results have indicated that under no load conditions, flux distribution in optimized stator and rotor cores has slightly been improved with respect to that in conventional stator and rotor cores, in particular, the magnetic flux density in the air gap.