50 kJ SMES magnet design optimization using real coded genetic algorithm

Abstract

SMES systems are being made and utilized for various applications i.e. dampening low frequency oscillations, instantaneous voltage drop compensation, pulsating power supply etc. Due to tape like structure of 2G HTS and fabrication limits of long continuous length , SMES coil is made via assembling individual single pancake coil/double pancaked coil in the solenoid or toroid fashion. A SMES magnet is designed optimally, using real coded genetic algorithm, with minimum length as objective function for fxed amount of stored energy(50 kJ), using 2G HTS tape. Simulated binary crossover and polynomial mutation function are used for crossover and mutation respectively. Elitism is also used for faster convergence of genetic algorithm. An minimum length is found for 50 kJ stored energy. Stress calculation is carried out for this confguration.