Conceptual Design and Optimization of a UAV Using Simulated Annealing

This paper provides details of an existing methodology for conceptual sizing and optimization of a UAV using Simulated Annealing optimizer. An existing methodology applicable for a baseline configuration of a UAV with a trapezoidal wing and conventional horizontal and vertical tail mounted on a fuselage, with a single turbofan engine is described. This methodology uses empirical formulae for estimation of aerodynamic parameters, mass breakdown, payload packaging and estimation of performance parameters and fuel required for flying various legs of a user-specified mission. In the present study, this methodology was critically reviewed and enhanced by modifying various modules. The sizing problem was posed in an optimization framework, with a total of ten design variables involving five parameters related to wing geometry, three parameters related to fuselage geometry, mission fuel mass, and engine thrust at sea-level static condition. Nine constraints are included; four related to the volume and geometry of the configuration, four related to desired point performance, and one from aeroelasticity considerations. The sizing methodology was validated against published results and then coupled to a Simulated Annealing evolutionary algorithm to obtain the optimum results for minimum all-up weight of the UAV. Feasible solutions with 11.2% lower all-up weight were obtained. A sensitivity analysis of various constraints revealed that relaxing the aeroelasticity constraints does not result in significant reduction in all up weight, while relaxing the constraint on sustained load factor resulted in 12% reduction on all-up weight.