The Semantic Asteroids
- Paper number
IAC-08.C1.2.2
- Author
Dr. Dario Izzo, European Space Agency (ESA), The Netherlands
- Coauthor
Dr. Tamas Vinko, The Netherlands
- Coauthor
Mr. Marco del Rey Zapatero, The Netherlands
- Year
2008
- Abstract
In the field of interplanetary trajectory design, engineers have always been accostumed to deal with the orbits of the few solar system objects that were thought to be of interest to space science. The planets, the known comets and the asteroid belt have always been the favoured targets of space missions, even in their conceptual phases. Only recently, our solar system found itself populated by thousands of smaller objects that were previously unseen: the Near Earth Asteroids. Thousands of different orbit geometries and phasings that represent a fascinating challenge to those who know about spacecraft trajectory optimisation. When speaking about common objects, such as Jupiter, any system engineer or trajectory designer is able to quote possible fly-by strategies, Δ V estimations, transfer time estimations, launch windows ... but when it comes to asteroids they are simply too many! Fortunately, recent advances in the automation of interplanetary trajectory optimisation makes it possible to compute a previously unthinkable number of trajectories by ]lq just pressing a button. Unfortunately, no common knowledge is usually extracted by this great number of computations that mission analysis experts make every month in different parts of the world.
In this paper we discuss about automated trajectory optimisation (low-thrust and chemical) to asteroids and, in particular, we describe the new computational algorithms in use at the Advanced Concepts Team of the European Space Agency that allow for a fully automated trajectory optimisation process. In particular, global optimisation techniques (Simulated Annealing, Differential Evolution, Particle Swarm Optimisation, Genetic Algorithm) are used cooperatively to optimise chemical propulsion missions and a new multiphase direct optimiser based on AMPL (and thus making use of analytical gradient information) is used for low-thrust trajectories. Both these methods are completely automated and can be called by other routines/scripts when needed. The final automated trajectory optimisation architecture relies on python scripts a) interpreting the user request, b)writing the code for the objective function to be optimised, c) invoking the chosen solver, d) interpreting the results.
We then discuss about the use of semantic web technologies to store and visualise the enormous amount of data can can be generated by letting the computers optimise trajectories autonomously. The semantic web is an XML markup language (based on the so called rdf triplets) that allow for a unique representation of relations between database entries. A first database of globally optimal trajectories to asteroids, named The Semantic Asteroids, has been preliminarly populated and made available to the scientific community via the web (www.esa.int/gsp/ACT/inf/op/SemanticAsteroids/TheSemanticAsteroids.htm). In our vision, the future system engineer will be able to go to the web and query a database of optimised trajectories. Should his particular transfer not be included in the database, a server will immediately start the automated computation and, when finished, store it into the database as to make the result available for possible future queries for similar trajectories. Our database is just a preliminary step towards this vision, but interestingly enough it shows how a very simple ontology (from wikipedia: in both computer science and information science, an ontology is a representation of a set of concepts within a domain and the relationships between those concepts. It is used to reason about the properties of that domain, and may be used to define the domain) allows for a quite complex trajectory query system.
- Abstract document
- Manuscript document
(absent)