Optimal Project of a Salient-Pole Synchronous Generator using Reluctance Networks

This work approaches the optimal design of a synchronous generator with the objective of reducing the active material mass. The optimization of a salient pole synchronous generator must be approached considering the generator armature reaction to meet the specifications of electrical power, terminal voltage and field density specified in the project. The armature reaction calculation made use of two reluctance networks that calculate the direct axis and quadrature axis inductances. The optimized generator has a smaller mass of active material, reducing its production cost and increasing its commercial competitiveness. To ensure the desired performance, electrical and magnetic constraints were added to the optimization problem. The networks adapt their reluctance values and flows in the sources, as the machine dimensions are adjusted by an optimization algorithm, Particle Swarm Optimization (PSO), which minimizes the generator mass along the iterations. The results were validated through computational simulations with the 2D finite element method.