Complex Science for a Complex World
Next
Complex Science for a Complex World
Table of Contents
Foreword
Acknowledgments
Sponsors
1. Building a New Science for a Complex World
1. Complex Science for a Complex World: An Introduction
Introduction
The complex world of human ecosystems
Understanding human behaviour
Complexity of human ecosystems
Exploring human ecosystems with agents
Autonomous and adaptive agents
Hierarchies of autonomous agents
Interconnected and embedded agents
Exploring national research priorities with agents
An environmentally sustainable Australia
Promoting and maintaining good health
Safeguarding Australia
Building a new science for a complex world
Towards holistic visions and polymorphic tools
Towards a new epistemology of science
A few last words
References
2. Towards a New Ontology of Complexity Science
Introduction
A fully connected world
A new kind of science
The practice of complexity
References
3. Agents, Icons and Idols
Abstract
Introduction
Cognitive agents
Interaction system
System of representation
Conative system
Organisation system
Exploring human ecosystems
Deceptive idols
Bacon’s idols
Prospect theory
Cognitive dissonance
Socially constructed icons
Peirce’s icons
Enactive cognitive theory
Social constructions
Paradigm shift
Normal views on social rationality
A post-normal temptation
Companion modelling
Conclusion
Acknowledgement
References
4. The Uncertain Fate of Self-Defeating Systems
Abstract
Introduction
Self-referential systems
Definition
The bar problem defined
The bar problem simulated
The bar problem extended
Human ecosystems
Fisheries
Evolutionarily stable strategies
A modified minority game?
Discussion
Acknowledgments
References
5. The Structure of Social Networks
Abstract
Introduction
Networks
Network properties
Real world complex networks
Models of complex networks
Hierarchies and dominance
Dominance and linear hierarchical ranking
Bobby Fischer and the Ruy Lopez opening line
Fischer’s dominance
Some closing comments on dominance hierarchies
Enforcement of social norms
Models of social dilemmas
Model of social structure
Discussion and implications
Some comments on the enforcement of social norms
Closing comments
Acknowledgements
References
6. Integration and Implementation Sciences: Building a New Specialisation
Abstract
Introduction
Defining key elements of the social issues
Theoretical and methodological pillars
Systems thinking and complexity science
Participatory methods
Knowledge management, exchange and implementation
Grounding in practical collaboration
Where would Integration and Implementation Sciences sit in universities?
Statistics as a useful analogy
Challenges in developing a new specialisation
Next Steps
Acknowledgments
References
2. Exploring National Research Priorities with Agents
Viewpoint from a Practitioner
7. Sustainability Assessment of Housing Developments: A New Methodology
Abstract
Introduction
Methodology
Definition of housing development systems
Housing development system models
Sustainability indicators and assessment
Effect of human behaviour
Obtaining behavioural information
Multi-agent system modelling
Case study application
Background
Christie Walk system and multi-agent representation
Defining behaviour
Multi-agent model implementation
Results of sustainability assessment simulations
Scenario analyses
Conclusion and future directions
Acknowledgements
References
Viewpoint from a Defence Expert
8. WISDOM-II: A Network Centric Model for Warfare
Abstract
Introduction
Limitations of existing agent-based distillation systems
Network Centric Multi-Agent Architecture
WISDOM Version II
Architecture
Characteristics of agents
Personalities of agents
Decision-making mechanism
Recovery
Visualisation
Reasoning
Terrain feature
Interactive simulation
Scenario analysis
Conclusion
Acknowledgement
References
Viewpoint from a Policy-Maker
9. Managing Agricultural Pest and Disease Incursions: An Application of Agent-Based Modelling
Abstract
Introduction
Exotic incursion management model
Karnal bunt case study
Karnal bunt of wheat
Case study region
Customised model
Farm Component
Quarantine Component
Regional Component
Scenario-based results
Incursion scenarios
Contractor based incursion
Fertiliser based incursion
Incentives to self report
Government policy implications
Conclusion and perspective
References
Viewpoint from a Policy Adviser
10. SimDrug: A Multi-Agent System Tackling the Complexity of Illicit Drug Markets in Australia
Abstract
Introduction
Complexity of illicit drug use and markets
The Australian heroin drought
Complexity Theory and illicit drug use
SimDrug—model description
Background
Advocating for a Multi-Agent System
Model description
UML Structure
Class diagram
Modelling sequence
Activity diagrams
SimDrug—preliminary results
Results
Sensitivity Analysis
Simulating the heroin drought
Conclusion
Acknowledgments
References
View from an Energy Expert
11. NEMSIM: Finding Ways to Reduce Greenhouse Gas Emissions Using Multi-Agent Electricity Modelling
Abstract
Introduction
Australia’s national electricity market
Operating the NEM
Electricity markets are complex adaptive systems
Approaches to electricity market modelling
Cournot equilibrium models
Supply Function Equilibrium Models
Conjectural variations approaches
Discussion
Simulation models
Agent-based models of electricity markets
A brief review of earlier agent-based models
Agent behaviour in Australia’s national electricity market
NEMSIM: the National Electricity Market simulator
Learning from experience
NEMSIM environments
NEMSIM as a greenhouse gas emissions calculator
Simulated outcomes
References
Viewpoint from a Regional Adviser
12. AtollGame: A Companion Modelling Experience in the Pacific
Abstract
Introduction
Collective knowledge
Theoretical assumptions
Global targeted appraisal
Individual activities survey
Processing elicited knowledge
Results from the GTA
Results from the IAS
AtollGame: the model
Elements of design
The virtual landscape
Water balance and hydro-geological models
Social entities
AtollGame: the role-playing game
Conceptual framework
The game
Playing rounds
Outcomes from the role-playing game
Exploring scenarios
Discussion and perspectives
Instant debriefing
Distant debriefing
References
View from a Principal Scientist
13. Multiple-Use Management Strategy Evaluation for Coastal Marine Ecosystems Using
InVitro
Abstract
Introduction
Methods
InVitro
Environments
Ecology
Fisheries
Other human sectors
Management of the Australian NWS regional ecosystem
Results
Fisheries management
Contamination
Coastal development—capacity and shipping traffic
Conservation implications
Conclusions
References
Viewpoint from a Practitioner
14. Rangeland Consolidation Patterns in Australia: An Agent-Based Modelling Approach
Abstract
Introduction
Methods
Pastoral enterprise submodel
Biophysical submodel
Results
Reference conditions
Sensitivity analysis
Discussion
Conclusions
References
List of Contributors
Index
List of Figures
3.1.
Belief-Desire-Intention (BDI) structure of a conative system
3.2.
Passive and active perception systems. (Adapted from Ferber 1999)
3.3.
Conative system and its two sub-systems. (Adapted from: Ferber, 1999)
3.4.
Structure of a Consumat agent
3.5.
Traditional cycle of interactions during a companion modelling process
4.1.
A simulated 100-week record of attendance at El Farol
5.1.
Random graphs
5.2.
Example of criticality phenomena in the evolution of graphs
5.3.
Examples of complex networks
5.4.
Erdös-Rényi model of random graph evolution
5.5.
Different subgraphs appear at varying threshold probabilities in a random graph (After Albert and Barabási 2002)
5.6.
Progressive transition between regular and random graphs
5.7.
The Ruy Lopez opening line
5.8.
Bobby Fischer’s network of immediate opponents
5.9.
Prisoners' Dilemma
5.10.
The architecture of the Norms and Meta-norms games (After Axelrod 1986)
5.11.
Simulation results
6.1.
The relationship between the home base and the key sectors for Integration and Implementation Sciences
7.1.
Housing development assessment methodology
7.2.
The complex housing development system
7.3.
Framework of interrelated models for housing development sustainability assessment
7.4.
The Sustainability Scale
7.5.
Cumulative distribution of mains water use exceeding the sustainability threshold level
7.6.
Mains water use frequency distribution and the sustainability threshold level
7.7.
Multi-agent representation of a housing development
7.8.
Christie Walk water model conceptualisation
7.9.
Christie Walk CO
2
model conceptualisation
7.10.
Christie Walk waste model conceptualisation
7.11.
Christie Walk economic model conceptualisation
7.12.
Christie Walk social model conceptualisation
7.13.
Occupant behaviour categories for total waste production
7.14.
Architectural plan of the Christie Walk development and the CORMAS model environment representation
7.15.
Sustainability Scale Ratings for Christie Walk for the 5 indicators
7.16.
Effect of waste production behaviour on waste sustainability
7.17.
Effect of recycling behaviour on waste sustainability
7.18.
Effect of water use behaviour on water sustainability
7.19.
Effect of energy use behaviour on CO
2
sustainability
8.1.
A coarse-grained view of NCMAA
8.2.
wo-layer architecture in NCMAA
8.3.
Influence Network for NCMAA in Warfare
8.4.
Command and Control Hierarchy in WISDOM-II
8.5.
Influence directions
8.6.
Initial position
8.7.
Damage of each force over time
8.8.
Average degree and average path length of the blue and red communication network over time
8.9.
Correlation coefficient between red damage and the knowledge correctness at agent level (left) and force level (right) (window size is 5)
9.1.
Structure of the EIM model
9.2.
Spatial visualisation of the environment: Quarantine response
9.3.
Area infested, contractor scenario
9.4.
Area infested, fertiliser scenario
9.5.
Receipts per infested farm, contractor scenario
9.6.
Receipts per infested farm, fertiliser scenario
10.1.
SimDrug Class Diagram (designed with VisualParadigm
©
)
10.2.
‘assessNeed’ Activity Diagram (designed with VisualParadigm
©
)
10.3.
‘useDrug’ Activity Diagram (designed with VisualParadigm
©
)
10.4.
‘declareOverdose’ Activity Diagram (designed wtih VisualParadigm
©
)
10.5.
Simulated Total and Fatal Overdoses over the 4-year period
10.6.
Evolution of hot spot locations over time from initial condition (left) to final state (right)
10.7.
Influence of outreach workers on overdose rates (left) and treated user rates (right)
10.8.
Influence of constables on number of arrested dealers (left) and maximum dealer's cash (right)
10.9.
Number of users under treatment according to increasing values for wealth updating rate
10.10.
Number of fatal overdoses derived from the base-scenario and from real data
11.1.
Inter-connectors in Australia’s National Electricity Market
11.2.
The NEM as a Complex Adaptive System
11.3.
Some trends in electricity market modelling
11.4.
Electricity prices in California, 1998-2000
11.5.
An overview of NEMSIM
11.6.
NEMSIM GHG emissions example windows for a generator plant: tabular form (top); graphical form (bottom)
11.7.
NEMSIM regional summary window for GHG emissions
12.1.
Tarawa Atoll (Bonriki and Buota islands are on the lower right of the atoll)
12.2.
Elder man in Abatao commenting on photos of economic activities
12.3.
Example of Card Game’s flowchart
12.4.
Partial view over an associative network
12.5.
Overall table ranking most quoted elements during the card game
12.6.
UML-based Class Diagram representation of the common ontology
12.7.
AtollGame environment
12.8.
Representation of the freshwater lens in AtollGame
12.9.
Island 1 (left) and Island 2 (right)
12.10.
Flowchart of financial, technical and social solutions
13.1.
The MSE Framework
13.2.
Estimated and actual stock sizes for the primary fishery target group (Lutjanids)
13.3.
Relative levels of contaminant in prawns within 20 km of an outfall under the status quo and enhanced management strategies
13.4.
Profit relative to profit at historical capacity for the port of Dampier under the range of port capacity levels allowed by the status quo and enhanced management strategies
14.1.
Results of reference treatment in a heterogeneous rangeland system
14.2.
Number of block sales for the reference treatment (solid line)
14.3.
Consolidated property size distribution at the end of the simulations using reference conditions
14.4.
Average consolidated property size response to the variable costs
14.5.
Mean consolidated enterprise size at the start and end of 113-year simulations
14.6.
Location of sites
List of Tables
1.1.
Agents simulated in Chapters 7-14
3.1.
Partial list of Cognitons proposed by Ferber
4.1.
Strategies and payoffs for homogeneous fleets
4.2.
Fleet strategies and payoffs when information is shared
5.1.
Example of ranking between players according to David’s score
5.2.
Top 10 players in Fischer’s gaming network
7.1.
Model sustainability indicators
8.1.
The set of hyper cells in between
8.2.
Scenario settings
9.1.
Regional economy effect of alternative incursion scenarios
10.1.
Influence of tested values on selected variables
11.1.
XML fragment with Generation Technologies description