Design optimization of a capacitor self-excited cage induction generator

Abstract

The design of a self-excited cage induction generator (SEIG) is formulated as a non-linear multivariable optimization problem. A set of nine variables is identified and nine constraints are imposed on the design to meet the desired performance of the machine. A set of objective functions, such as the active material cost, capacitor cost, regulator cost, and the sum of the active material cost and regulator cost, is considered for the design optimization of the SEIG. Each objective function in the design is optimized using Rosenbrock's method of direct search in conjunction with the sequential constrained minimization technique (SUMT). The computed design results for the various objective functions for a 3.7 kW self-excited cage induction generator are presented in tabular form.