Use of Information Theory Techniques with System Dynamics Models

Abstract

Research was done to show how the output distribution of a system with random variables changes as the input distributions and other parameters are changed. Using Information Theory techniques, these different input and output distributions were analytically studied to try to quantify the uncertainty that exists in both the input and output distributions and to determine any relationships between the input uncertainty and output uncertainty. A simplified System Dynamics model of a generic Nuclear Power Plant Construction project was then created. The model was simple enough to allow for relative ease in analyzing data, but detailed enough to allow for changes within the model that could be interpreted as policy changes within the organization. The model was then run with various distributions for different input variables. Data was collected from these simulations for analysis. Using the results of the analytic research mentioned above, this data was analyzed to determine the feasibility of using Information Theory to enhance System Dynamics as a tool to help organizations with policy-making decisions. Results show that use of Information Theory to complement System Dynamics would be helpful in such situations.