Collective Intelligence for Control of Distributed Dynamical Systems

Collective Intelligence for Control of Distributed Dynamical Systems
Europhysics Letters , Vol. 49, No. 6, March 2000. (Postscript - 350KB )

Abstract: We consider the El Farol bar problem (W. B. Arthur, The American Economic Review , 84(2): 406--411 (1994), D. Challet and Y.C. Zhang, Physica A , 256:514 (1998)) as an instance of the general problem of how to automatically configure a distributed dynamical system so that its nodal elements do not ``work at cross purposes'', in that their collective dynamical behavior successfully achieves a global goal. We present a summary of a mathematical theory for such automated configuration applicable when (as in the bar problem) the global goal can be expressed as minimizing a global energy function. We then investigate the applicability of the core concepts of that theory to a difficult variant of the bar problem. We show that a system designed in accord with that theory performs nearly optimally for the bar problem, and in particular avoids the tragedy of the commons for that problem.