Information Related to Ongoing Research to
Develop a Decision Support System for
Trumpeter Swan Management
 
 
 
 
 
 
 

This is a collaborative project between the Northern Rocky Mountain Science Center and Midcontinent Ecological Science Center of the USDI-Geological Survey, and the Migratory Bird Program of the USDI-Fish and Wildlife Service (Region 6), and is jointly funded.

• Explore the "Breeding Habitat Needs for Trumpeter Swans" knowledge base (in development)

• Explore the "Management of Palustrine Wetlands in the Northern Rockies" knowledge base (in development)

• Explore the "Contribution an Area Can Make Towards the Flyway Management Plan" knowledge base (in development)

Notes on using the DECAF agent architecture for developing the decision support system:

• view the DECAF plans representing each agent in the multiagent approach for trumpeter swan management

• information about DECAF (Distributed, Environment-Centered Agent Framework), a project coordinated by Dr. Keith S. Decker, Department of Computer and Information Sciences, University of Delaware

Notes from the knowledge engineering sessions:

• the breeding habitat knowledge base
• Documentation for the ecological logic behind the rules in the breeding habitat knowledge base
• Flowchart depicting the general modular tree structure of the knowledge base, "Breeding Habitat Needs for Trumpeter Swans", which currently consists of 157 rules
• the migration chronologies and pathways knowledge base
• Tables delineating probabilities of migration linkage
• the montane wetland management knowledge base
• Flowchart depicting the general modular tree structure of the knowledge base, "Management of Palustrine Emergent Wetlands in the Northern Rockies", which currently consists of 1,852 rules
• the principles of flyway management knowledge base
• Flowchart depicting the decision tree entitled, "Can this area contribute to the RMP Plan's breeding population goals?"
• Flowchart depicting the decision tree entitled, "Can this area contribute to the RMP Plan's objective of redistributing wintering swans?"
• integrating distributed AI and flyway management
• Table delineating consequences of recommended water levels on wetlands and swans

Notes about the queuing model:
• base algorithm for the distribution simulator using probabilities of movement based on expert opinion
• example of output from the distribution simulator using the base algorithm only --NOTE: this is provided ONLYas an example, and does not represent completed output from the decision support system.
• draft algorithm to redistribute swans that adjusts probabilites of movement by incorporating information on wetland condition and principles of flyway management
• time line delineating definitions related to seasons for the queuing model
• documentation for distances between trumpeter swan September survey areas
Sojda, R. S., and A. E. Howe.  1999. Applying cooperative distributed problem solving methods to trumpeter swan management.  Pages 63-67 in: Environmental Decision Support Systems and Artificial Intelligence, Papers from the American Association for Artificial Intelligence Workshop: Technical Report WS-99-07.  AAAI Press.  Menlo Park, CA.  [postscript format]

Slides from Sojda and Howe's paper from the Environmental Decision Support Systems and Artificial Intelligence Workshop, sponsored by the American Association for Artificial Intelligence (AAAI-99)

Potential rule bases and frames

Example of low level rules

About modularity in rule bases

Draft discussion document about using multiagent systems and BDI architectures
 

...in collaboration with the Environmental Statistics Group in the Biology Department at Montana State University

Notes from conference call (8 November 1999) on data analysis of swan numbers

Links to more information

Detailed study plan for this project

General description of this project

Rick Sojda's web page