A Hybrid Algorithm by Combination of Genetic Algorithm and Local Optimization for Constrained MHD Pump Optimization

The pumping of liquid metals by magnetohydrodynamic (MHD) converters, without any moving parts are competing with conventional hydraulic pumps. Thus, the movement of the fluid is a consequence of the interaction the magnetic and electric fields. This work relates to the application of an approach for optimal design of a DC magneto-hydrodynamic pump. The optimization method is based on the solution of non-linear constrained problem by combination global and local optimization method. This paper applies the hybrid algorithm by combination of genetic algorithm and Fmincon for optimal design of DC MHD pump. The comparative study of the obtained results shows that the performance of the hybrid method is effectively improved.