The flux switching machine is a relatively new class of reluctance machine in which power is converted mainly by changing mutual coupling between windings. This thesis develops the parameterised finite element analysis of this machine and introduces the 'flux map', an extension to the flux-mmf-position electromagnetic characterisation approach to fully account for mutual coupling between two phases. A dynamic simulation model of the flux switching machine and the flux switching motor including its drive and control are then developed based on this data. Comparison of experimental and simulated results shows good agreement in both cases and the effects of various factors on simulated output are examined. A genetic algorithm system is integrated with parameterised finite element analysis and its operation demonstrated for design cases using flux switching and switched reluctance machines and it is shown that the system is applicable throughout the design and optimisation process. A simple image processing method for iron loss visualisation in electromagnetic devices is also developed and demonstrated for the flux switching motor. The flux map is applicable to any reluctance machine where mutual coupling between two phases must be accounted for while the genetic algorithm system and iron loss visualisation technique may be applied to virtually any electromagnetic system which can be modelled in 2-dimensional finite element analysis.