Interests in using rare-earth free motors such as switched reluctance motors (SRMs) for electric and hybrid electric vehicles (EV/HEVs) continue to gain popularity, owing to their low cost and robustness. Optimal design of an SRM, to meet specific characteristics for an application, should involve simultaneous optimization of the motor geometry and control in order to achieve the highest performance with the lowest cost. This dissertation firstly presents a constrained multi-objective optimization framework for design and control of a SRM based on a non-dominated sorting genetic algorithm II (NSGA-II). The proposed methodology optimizes SRM operation for high volume traction applications by considering multiple criteria including efficiency, average torque, and torque ripple. Several constraints are defined by the application considered, such as the motor stack length, minimum desired efficiency, etc. The outcome of this optimization includes an optimal geometry, outlining variables such as air gap length, rotor inner diameter, stator pole arc angle, etc as well as optimal turn-on and turn-off firing angles. Then the machine is manufactured according to the obtained optimal specifications. Finite element analysis (FEA) and experimental results are provided to validate the theoretical findings. A solution for exploring optimal firing angles of nonlinear current-controlled SRMs is proposed in order to minimize the torque ripple. Motor torque ripple for a certain electrical load requirement is minimized using a surrogate-based optimization of firing angles by adjusting the motor geometry, reference current, rotor speed and dc bus voltage. Surrogate-based optimization is facilitated via Neural Networks (NN) which are regression tools capable of learning complex multi-variate functions. Flux and torque of the nonlinear SRM is learned as a function of input parameters, and consequently the computation time of design, which is crucial in any micro controller unit, is expedited by replacing the look-up tables of flux and torque with the surrogate NN model. This dissertation then proposes a framework for the design and analysis of a coreless permanent magnet (PM) machine for a 100 kWh shaft-less high strength steel flywheel energy storage system (SHFES). The PM motor/generator is designed to meet the required specs in terms of torque-speed and power-speed characteristics given by the application. The design challenges of a motor/generator for this architecture include: the poor flux paths due to a large scale solid carbon steel rotor and zero-thermal convection of the airgap due to operation of the machine in vacuum. Magnetic flux in this architecture tends to be 3-D rather than constrained due to lack of core in the stator. In order to tackle these challenges, several other parameters such as a proper number of magnets and slots combination, number of turns in each coil, magnets with high saturated flux density and magnets size are carefully considered in the proposed design framework. Magnetic levitation allows the use of a coreless stator that is placed on a supporting structure. The proposed PM motor/generator comprehensive geometry, electromagnetic and mechanical dimensioning are followed by detailed 3-D FEA. The torque, power, and speed determined by the FEA electromagnetic analysis are met by the application design requirements and constraints for both the charging and discharging modes of operation. Finally, the motor/generator static thermal analysis is discussed in order to validate the proposed cooling system functionality.
