Skip to Main content Skip to Navigation

Compilation de connaissances pour la décision en ligne : application à la conduite de systèmes autonomes

Abstract : Controlling autonomous systems requires to make decisions depending on current observations and objectives. This involves some tasks that must be executed online--with the embedded computational power only. However, these tasks are generally combinatory; their computation is long and requires a lot of memory space. Entirely executing them online thus compromises the system's reactivity. But entirely executing them offline, by anticipating every possible situation, can lead to a result too large to be embedded. A tradeoff can be provided by knowledge compilation techniques, which shift as much as possible of the computational effort before the system's launching. These techniques consists in a translation of a problem into some language, obtaining a compiled form of the problem, which is both easy to solve and as compact as possible. The translation step can be very long, but it is only executed once, and offline. There are numerous target compilation languages, among which the language of binary decision diagrams (BDDs), which have been successfully used in various domains of artificial intelligence, such as model-checking, configuration, or planning. The objective of the thesis was to study how knowledge compilation could be applied to the control of autonomous systems. We focused on realistic planning problems, which often involve variables with continuous domains or large enumerated domains (such as time or memory space). We oriented our work towards the search for target compilation languages expressive enough to represent such problems. In a first part of the thesis, we present various aspects of knowledge compilation, as well as a state of the art of the application of compilation to planning. In a second part, we extend the BDD framework to real and enumerated variables, defining the interval automata (IAs) target language. We draw the compilation map of IAs and of some restrictions of IAs, that is, their succinctness properties and their efficiency with respect to elementary operations. We describe methods for compiling into IAs problems that are represented as continuous constraint networks. In a third part, we define the target language of set-labeled diagrams (SDs), another generalization of BDDs allowing the representation of discretized IAs. We draw the compilation map of SDs and of some restrictions of SDs, and describe a method for compiling into SDs problems expressed as discrete continuous networks. We experimentally show that using IAs and SDs for controlling autonomous systems is promising.
Document type :
Complete list of metadatas
Contributor : Alexandre Niveau <>
Submitted on : Wednesday, November 28, 2012 - 4:55:18 PM
Last modification on : Monday, October 19, 2020 - 11:07:10 AM
Long-term archiving on: : Saturday, December 17, 2016 - 3:58:22 PM


  • HAL Id : tel-00758266, version 1



Alexandre Niveau. Compilation de connaissances pour la décision en ligne : application à la conduite de systèmes autonomes. Intelligence artificielle [cs.AI]. Université Paul Sabatier - Toulouse III, 2012. Français. ⟨tel-00758266⟩



Record views


Files downloads