Maffione Porzia Federica
Politecnico di Torino
Italy
Title: Sequential quadratic programming and metaheuristic methods applied to low-thrust interplanetary transfers
Biography
Biography: Maffione Porzia Federica
Abstract
Using transcription methods, the trajectory optimization problem can be easily converted into a non-linear programming problem. Once a set of variables and constraints are defined, one of the most efficient methods to optimize the objective function is the sequential quadratic programming.
In this work, a tool that combine the SQP with two different types of metaheuristic methods (the particle swarm optimization and the differential evolution) has been developed. The results obtained with the SQP were validated using the MATLAB software FALCON.m developed at the Institute of Flight System Dynamics of Technische Universität of München. Since the PSO and the DE require converting the constrained optimization problem into an unconstrained one, a penalty function with static weights is also provided.
In this work, low-thrust interplanetary transfers with electric propulsion are analyzed. The strong correlation between the weights of the penalty function, the number of generations and the accuracy of the solution is shown up with a large variety of examples. Then, the results will show if there is a real advantage using the combined method in terms of computational time and objective function.
