| The ability to simulate, understand and ultimately control complex and living systems is a field in its infancy. Currently, there are few adequate models and even fewer simulation tools. The simulation tools that do exist are primitive, both functionally and visually, and offer very little real-time feedback. Knowledge about how such systems self-organize and behave as well as prediction of system behavior is lacking, in part, due to the absence of these models and simulations.; This research included the expansion and modification of the classic Holland ECHO models by the addition of new models, to better simulate natural systems for the purpose of pinpointing key characteristics of complex adaptive systems (CAS). A new model that I originated was “resource hierarchies”, in which higher-level resources are composed of lower-level resources, intended at better simulating molecular organic processes found in nature. A prime goal of the simulation tool was to construct a general purpose software model of a CAS that exhibits processes and dynamics similar to real-life systems, with the ability to rerun simulations from different initial states. An additional goal of implementing this simulation tool was to create this system with support for real-time 3D visualizations, allowing CAS researchers the ability to observe emergent processes in a way that was previously unavailable.; As a result of these goals, I have created an implementation and visualization based on John Holland's ECHO models using a high-speed 3D graphics engine based on OpenGL. The simulation software that I have developed, ECHOSTAR (ECHO Simulation Tool and Renderer) allows experimenters the ability to witness the evolution of CAS.; Finally, I have interpreted some of the first output tables from ECHOSTAR in an attempt to identify key relationships between adaptive agents, limited resources, and their environments. The ECHOSTAR software's visualizations and data logs provide a means by which to easily identify and experiment with CAS behavior.; I hope that the implementation and use of this simulation software, which allows researchers to observe the development of CAS over time and dissect the results, will lead to better understanding of all CAS. |
