Application of optimization algorithms in the design of a superconducting A.C. generator rotor

The superconducting a.c. generator is expected to be the optimum choice among a.c. generation systems in future because of its reduced size, high efficiency, high terminal voltage and its contribution to the stability of the power system. Such machines also exhibit unique design problems which remain unsolved. The optimal selection of the basic design parameters is a current problem of interest. This thesis is intended as a contribution in this direction, and a general design strategy has been developed for the superconducting a.c. generator. Elements of the design process include magnetic field analysis, losses, and mechanical performance all which of are discussed in the thesis. An analytical model has been developed to help determine the distribution of magnetic flux density inside the superconducting machine. This model takes into account the number, and the geometric structure, of the winding slots and allows the rotor of the superconducting machine to be designed with optimum magnetic field distribution. A general design strategy has been developed for the superconducting a.c. generator rotor for predicting the optimum design. The design optimization process incorporates "direct search" and random-shrinkage methods. Two direct search methods of minimization have been compared on mathematical functions and also on machine design problems. The best method is highlighted and discussed. A general computer program package is presented that will optimize and analyse machine design problems. The package is organised in such away that future addition or deletion of performance specifications, constraints, optimization methods and design process elements are readily implemented.