1 / 16

Stéphane Vialle Stephane.Vialle@supelec.fr Eugen Dedu Eugen.Dedu@prism.uvsq.fr

Charles Hermite Center (France). ERSIDP Supélec (France). ParCeL-5/ParSSAP: A Parallel Programming Model and Library for Easy Development and Fast Execution of Simulations of Situated Multi-Agent Systems. SNPD’02. Stéphane Vialle Stephane.Vialle@supelec.fr

jalene
Télécharger la présentation

Stéphane Vialle Stephane.Vialle@supelec.fr Eugen Dedu Eugen.Dedu@prism.uvsq.fr

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. Charles Hermite Center (France) ERSIDP Supélec (France) ParCeL-5/ParSSAP: A Parallel Programming Model and Library for Easy Development and Fast Execution of Simulations of Situated Multi-Agent Systems SNPD’02 Stéphane VialleStephane.Vialle@supelec.fr Eugen Dedu Eugen.Dedu@prism.uvsq.fr Claude Timsit Claude.Timsit@prism.uvsq.fr

  2. ?? C-library …. 1 - Motivations 1 - Situated MAS with fine modeling: complex to program (ex: simultaneity of agent actions and conflict resolutions) 2 - Fine or large situated MAS: very time consuming (for simultaneity modeling, fair conflict resolution, …) 3 - MAS: natural parallelism different of computer parallelism (efficient adaptation is difficult) • We need: • Parallel programming models allowing easy developments • Parallel development tools on modern MIMD parallel computers

  3. 2D-Grid 2D-Torus 4-connectivity 8-connectivity 2 - ParCeL-5 Model: Basic Components 4 basic components :Environment, Resources, Agent/Cell, Arbitrator • 1 -The Environment : • 2D-Grid or 2D-Torus • (torus = circular grid) • 4-connectivity or • 8-connectivity • Only one agent per box • An obstacle entirely • fills a box

  4. Mines Factories 2 - ParCeL-5 Model: Basic Components 4 basic components : Environment, Resources, Agent/Cell, Arbitrator • 2 -The Resources : • Resources store or consume • some objects (ores) • Any number of resources • Any number of kind of resources • (any kind of ores) • Only one resource per box, • but an agent can still enter the box

  5. Local memory (dynamically extensible) Priority and Speed Percepts of the environment Behavior function later 2 - ParCeL-5 Model: Basic Components 4 basic components : Environment, Resources, Agent/Cell, Arbitrator • 3 -The Agents/Cells : • Local memory • Perception of their local • environment • Private behavior (static) • They are mobile agents : • Declared before running • Dynamically created • or destroyed : Agent skeleton

  6. Winner agent Conflict ! Arbitrator 2 - ParCeL-5 Model: Basic Components 4 basic components : Environment, Resources, Agent/Cell, Arbitrator • 4 -The Arbitrator : • A set of rules modeling the physical law of the environment • Automatically applied: • when agents attempt to act • when environment is updated • Manages all conflicts, • with fairness and • random choices

  7. Potential field detection Direct vision 2 - ParCeL-5 Model: Agent Behaviors Agent Behavior: • Agents: Perceive their local environment • Act on their local environment Agent Percepts: • Unlimited number of percepts • But any percept must be LOCAL • Direct vision • Potential field detection, …. Agent actions: • Actions on env.: move, take/drop objects, • Actions on agents: create/destroy other agents • LOCAL actions

  8. 2-ParCeL-5 Model: Cyclic Running Simulation beginning parallel sequential Parallel initialisation step (environment, resources, percepts) Step 1: Activation of agent behaviors, declaration of action intentions Step 2: Resolution of agent conflicts (by the arbitrator) Simulation loop Step 3: Execution of actions of winners and non-conflicting agents Step 4: Execution of parallel or sequential end-cycle function (user) Step 5: Update of the environment and agent percepts Simulation end  Mainly Parallel

  9. Agent run Conflict management Environment update Cyclic running … Parallelization … User ParCeL-5 model ParCeL-5 library Behavior functions 2-ParCeL-5 Model: Easy Situated MAS Developments

  10. Shared Memory PE-0 PE-1 PE-2 PE-3 3 - Parallelization: Strategy • Domain partitioning strategy • SPMD paradigm • Multithreading implementation on shared memory Classical Environment: domain partitioning Synchronization barriers • Optimized parallel and sequential design of : • percept management • environment update Specific and Optimized

  11. Agent following an increasing potential path 1 2 1 Agent out potential field 1 1 2 3 2 2 3 4 3 1 5 3 4 4 2 1 1 2 1 Obstacle 4 5 6 5 3 2 1 2 3 2 1 6 5 6 5 3 2 3 4 3 2 1 4 Strong source 7 7 4 5 6 5 4 3 2 3 4 4 3 2 1 5 Medium source 3 4 5 4 3 2 1 2 3 4 3 2 1 5 Potential field frontier 3 - Parallelization: Wave Potential Field Propagation • Potential propagation principles : • Potentials emitted by resources (function of their features) • Spread, decrease, and bypass obstacles • Detected by agents (to avoid obstacles and reach resources) • Fast and sufficient composition of potential fields (max( ) operator)

  12. Parallelization : • Automatic parallel use from agent behaviors • Explicit parallel update: • Domain decomposition • Local propagation • Frontier sharing ; Re-propagation ; Until no change Frontier Complex potential propagation P0 slice Potential field P1 slice Potential source Obstacle 3 - Parallelization: Wave Potential Field Propagation • Several sequential algorithms • Several complex combinations in parallel algorithm

  13. Breadth first recursive method: • most evolved, less efficiently implemented • re-propagation & numerous obstacles • Regular loop iterative method: • less evolved, most efficiently implemented • propagation • re-propagation & numerous potential sources 3 - Parallelization: Wave Potential Field Propagation Ex : 1024x1024 boxes, 10000 resources, 0 obstacle, SGI-Origin2000

  14. 1 - Carrier robot simulation : • 1024 x 1024 boxes, 100 cycles • 10000 – 100000 agents • Agents catch and carry some ores • Search efficient carrier population • S(32) = 15 (SGI) • S(4) = 2.7 (SGI) – 3.5 (SUN) Efficient parallel runs % of cooperative agents time 4 – Applications and Global Perfs • 2 – Game of life (Conway) : • Immobile agents … • Possible with ParCeL-5 ! • 3 - Society of competitive selfish agents : • Ishida – Yokoi – Kakazu, ECAL’99 • One agent : does nothing or cooperates = f(random, agent behaviors) • How many cooperate ? • Same results than original simulations • Correct situated MAS simulator Stabilization phenomena

  15. 5 - Conclusion and Future Works • Present results of ParCeL-5/ParSSAP: • Good perfs : on high cost SGI-Origin2000: S(32) = 15 • on small 4-processor SUN: S(4) = 3.5 • Allows quick developments of situated MAS • Runs simulations with more than 100000 agents • Open parallel programming model : vision, potential, … • In the future …. • Add direct communication between agents • Add heuristic choice of wave propagation algorithms • Continue to build large situated MAS for emergence search • Optimize for low cost multiprocessor PC

  16. ParCeL-5/ParSSAP: A Parallel Programming Model and Library for Easy Development and Fast Execution of Simulations of Situated Multi-Agent Systems Questions ?

More Related