Development Environments and Systems Architectures for Hybrid Agent-Stochastic Event Models of Socio-Technical Organizations

摘要: AbstractThis is a public version of the report delivered in February, 2000 from Los AlamosNational Laboratory team on Decision Structures (DS) project to Physical Science Lab-oratory (PSL) New Mexico State University (NMSU). It includes proposed simulationenvironment for hybrid agent stochastic-event modeling; experimental paramateri-zation models general; and consideration various platforms developmentof models, with special attention Swarm DEVS/HLA. Contents 1 Introduction 22 A Hybrid Simulation Environment DS Project 22.1 HybridSimulationEnvironments 22.2 SpecificArchitecture 53 Experimental Parameterization Agent Simulations 74 Development Environments 94.1 Requirements 94.2 PrimaryDevelopmentEnvironments 104.2.1 DEVS/HLA,UniversityofArizona[5] 104.2.2 Swarm,SantaFeInstitute[18] 134.2.3 ComparisonandRecommendation 134.3 SecondaryDevelopmentEnvironments 144.3.1 Mozart/OZ(Europeanacademicconsortium)[12] 144.3.2 AgentSheets[1] 144.3.3 The Paracell Programming (Flavors Corp.) [13] . 154.3.4 Strictly Declarative Modeling Language (SDML) (Centre Policy Model-ing,ManchesterMetropolitanU.)[17] 154.3.5 Ascape (Center Social Economic Dynamics, Brookings Institute)[3] 154.3.6 Repast (University Chicago’s Research Computing) [14] 164.3.7 Logic Ecological (LEM, Serguei Krivov, Jawaharlal Nehru Uni-versity)[11] 16