Reducing sub-harmonic spatial fields in Concentrated winding machines using Genetic Algorithms
Certain slot/pole combinations in fractional slot winding machines with concentrated coils exhibit high amount of sub-harmonic spatial fields. These harmonics may cause additional losses and ripple torque. Literature suggest solutions to reduce the amount of sub-harmonics by increasing the number of layers, varying the numbers of turns per coil or introducing flux barriers in the yokes of the machine.
This article presents solutions to minimize sub-harmonics in a 24 slot 22 pole fractional slot winding machine with concentrated coils. A Genetic Algorithm is used to optimize the design of multiple layer windings with different turns per coil and implementation of flux barriers in the yokes of the machine. The proposed solutions are verified with Finite Element Analysis and compared with respect to sub-harmonic content, losses and ripple torque.
Fourier transform of the analytical Magnetomotive Force produced by the winding is used to characterize the sub-harmonic content. A Genetic Algorithm optimizes the winding layout using multiple partial fitness functions. Both double layer and single layer winding constraints are implemented to construct different layouts.
Flux Barriers are implemented using Finite Element Analysis and Magnetic Equivalent Circuits to optimize the design and arrangement.
Calculations revealed a 85 % decrease in the 1.th sub-harmonic using the double layer windings proposed by the Genetic Algorithm, and a 46 % reduction in solid losses compared to the single layer case.
A four layer winding completely removed the sub-harmonic content in the machine, consequently decreasing the solid losses by 63 % compared to the single layer case.
By optimizing the winding layout it is possible to completely remove the sub-harmonics. However, this complicates the winding layout, thus application specifications will govern the solution chosen.
Flux barriers presents a simple and promising solution reducing sub-harmonics.
Place, publisher, year, edition, pages
Institutt for elkraftteknikk , 2013. , 42 p.
IdentifiersURN: urn:nbn:no:ntnu:diva-23604Local ID: ntnudaim:9553OAI: oai:DiVA.org:ntnu-23604DiVA: diva2:676811
Nilssen, Robert, ProfessorRøkke, Astrid