The optimization of flowfield of a self-field magnetoplasmadynamic (MPD) thruster was conducted by two soft-computing methods. Both the genetic algorithm (GA) and a new method to obtain a minimum/maximum based on the path integral were used to establish the optimum geometry that produces the highest thrust for specified operating conditions within a quasi-one-dimensional framework. The optimum geometry was found to be a quickly convergent and divergent geometry regardless of the method employed, and the optimized geometry was found to be the same that was obtained by the classical variational methods. This fact as well as their applicability to parallel computers assures that the soft computing methods are suitable for a more complicated multi-dimensional flowfield optimization problem.
