1 / 1

R. Stranders , A. Farinelli , A. Rogers, N. R. Jennings

Decentralised Coordination of Continuously Valued Control Parameters using the Max-Sum Algorithm. R. Stranders , A. Farinelli , A. Rogers, N. R. Jennings School of Electronics and Computer Science University of Southampton, UK. Motivation. Approach

oren
Télécharger la présentation

R. Stranders , A. Farinelli , A. Rogers, N. R. Jennings

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Decentralised Coordination of Continuously Valued Control Parameters using the Max-Sum Algorithm R. Stranders, A. Farinelli, A. Rogers, N. R. Jennings School of Electronics and Computer Science University of Southampton, UK Motivation Approach Max-Sum Algorithm used as starting point Availability of devices acquiring and processing information from the environment. Problem Formulation: global welfare maximisation through localcomputation: • Part of the GDL framework • Graphical models, Information theory • Message Content • Functions of variable states • Message Propagation • From Variable: aggregate information (Sum) • From Functions: maximisesum of utility and variable messages (Max) • Optimal on trees, good approximation on cycle graphs Variable nodes Function nodes • Goal: Coordinate the activities of a set of devices characterised by continuously valued control parameters. For example: Activation Time Heading and Velocity Desired Temperature From Discrete to Continuous Empirical Evaluation Main Technical Contribution Application Domain: Coordinate duty cycles of energy aware sensors to maximise event detection probability Representation: Continuous Piece-Wise Linear functions Max-marginalisation: Project Extract upper envelope Addition: Merge domains Add values • Main Issues: • Representation: how do we represent continuous utilities? • Operations: need to redefine main operations of max-sum: • Max-marginalisation • Addition References: Contacts: R. Stranders, A. Farinelli, A. Rogers, N. R. Jennings, School of Electronics and Computer Science University of Southampton Southampton, SO17 1BJ, UK. {rs06r,af2,acr,nrj}@ecs.soton.ac.uk • Stranders, R., Farinelli, A., Rogers, A. and Jennings, N. R. DecentralisedCoordination of Continuously Valued Control Parameters using the Max-Sum Algorithm, AAMAS-08 • Stranders, R., Farinelli, A., Rogers, A. and Jennings, N. R. Decentralised Coordination of Mobile Sensors Using the Max-Sum Algorithm, IJCAI-09

More Related