ICNPAA 2010: Mathematical Problems in
Engineering, Aerospace and Sciences

Integrated design of space launch vehicle is a complex and highly coupled problem and always demands well suited optimization methods to be employed that can span the design space to identify the most promising design. Non-dominated sorting based genetic algorithm has potential to optimize the multidisciplinary design of the entire vehicle at conceptual level. In this work, each constraint of the problem is treated as separate objective in addition to the original objective function. By using a multi-objective approach we aim to decrease the violation of constraints, but at the same time we look for objective function improvement, and this behavior may open up the potential of reaching the feasible region from different directions. The approach is demonstrated by application to the integrated design of a four stage solid launch vehicle. The mission is to deliver predefined payload to the low earth (circular) orbit. An existing, real world system has been used to validate the multistage launch vehicle system model.