Multidisciplinary optimization of a pump-fed system in a cryogenic LPE using a systematic approach based on genetic algorithm

The Optimal operation point of the system, and simultaneously the subsystems of a particular pump-fed system in a cryogenic liquid propellant engine are determined by a hybridized optimization method. Russian LPE RD-180 scheme is considered as the case study here. A parametric model from the different sub-systems of the case study, which considers their primary interrelationships, is firstly developed and the optimal operational design point of each subsystem is individually determined. These pre-optimal outcome ranges of subsystems parameters are used as the constraints of optimization method to find optimal operation point of the whole system. Fitness functions are a complex of engine system and subsystem parameters. Optimization method is a new method, which is a combination of the traditional gradient method and the conventional genetic algorithm. This method speeds up convergence approximately twice, and increases accuracy of choosing subsystems parameters in comparison with the traditional design procedure. The mean value of the fitness functions, after 70 iterations of this model, converges to the values, which are close to the prototype engine. The procedure of this paper could be used at low computational costs, as a reference for practical design of similar pump-fed systems.