Multiobjective Optimization of a Shell-Like Induction Spherical Motor for a Power-Assisted Wheelchair

In this paper, a new topology of a spherical machine, denominated as a shell-like spherical induction machine, which was initially studied in recent published papers, is now designed and optimized for a powered-wheel application. The main objective is to perform an optimization design for the power/volume ratio maximization. The multiobjective/multiconstraint evolutionary algorithm NSGA-II (elitist nondominated sorting genetic algorithm) was used as the optimization tool. This optimization tool provides us a first design and physical insight to be further used in more complex 3-D finite-element analysis (FEA) tools. Thus, the shell-like spherical induction machine electromagnetic and thermal analytical models are first developed. To validate the thermal model, a small-scale prototype was built using a new soft magnetic material as the stator and rotor magnetic core. Tests to the prototype were carried out in our laboratory, measuring its electromagnetic torque and temperature, comparing them with the FEA and analytical results. Then, the optimization problem is defined by describing the objective functions, decision variables, and constraints for the powered-wheel application. Finally, the optimization results are analyzed using the Pareto front approach, showing the evolution of each objective function and decision variable.