Optimisation Algorithm for Space Trajectory Problems
This project regards the development and testing of stochastic based algorithms for the global optimisation of interplanetary trajectories. The structure of this kind of problems has a multi-funnel shape with many thousands of local minima (see figure) and global optimisation techniques have to be used to find the global optimum.
In the last decade many researchers have used global optimization techniques to find optimal solutions to space trajectory design problems and many different methods have been proposed and tested on a variety of cases from pure genetic algorithms to evolutionary strategies to differential evolution to hybrid methods.
ASCL (Prof. M. Vasile) and cFASTT (Dr. E. Minisci) research groups recently proposed and tested a hybrid evolutionary algorithm which demonstrated to have optimal performance on a range of problems, when its main parameters are robustly tuned. Current work is devoted to the improvement of the algorithm in a way that the structure of the problem is detected and critical parameters of the algorithm are self-adapted during the optimisation. Since this kind of problem is paradigmatic of many other design problems, all developed algorithms have a tremendous potential for applicability within other engineering areas.
Figure 1. Cut of Rosetta mission search space. Each point is a local minimum.
For this task the HPC facility is fundamental for the testing part of the work, when statistics of performance on relevant test cases have to be produced. In fact, due to the stochastic nature of the optimisation algorithm, the robustness/reliability of the algorithm should be demonstrated demonstrated in statistical terms by multiple runs.
This research work is carried out in collaboration with Prof Massimiliano Vasile, Professor of Space Systems Engineering, Space ART – Advanced Space Concepts Laboratory, Department of Mechanical & Aerospace Engineering, University of Strathclyde. For more information about the project contact Dr Edmondo Minisci (email@example.com), Lecturer at the Department of Mechanical and Aerospace Engineering at the University of Strathclyde.