Multi-parameter optimization program for magnetostatic circuit design utilizing a genetic algorithm

This paper discusses the application of genetic algorithms to optimize a magnetostatic circuit problem. The use of a genetic algorithm, in conjunction with a standard finite element analysis software package, allows several design parameters to be optimized simultaneously. As an example, a magnetostatic circuit consisting of a linear magnet with a pole piece is optimized to minimize magnetic material volume, while maintaining at least a minimum field value. Basic genetic algorithm theory is discussed