Market-based multiagent fuzzy control

This paper introduces an approach to develop distributed fuzzy control systems within the framework of market-based control. In market-based systems an artificial economy is created defining the role of consumer and producer agents and a pricing system as well. The control problem is formulated as a distributed resource allocation problem in which prices act as coordination variables to induce market equilibrium. Here fuzzy rules are used to model the behavior of the consumer and producer agents. Agents play the role of subsystem controllers. Equilibrium is found by simultaneously solving the set of fuzzy models of the agents. An example concerning a set of coupled tanks is used to illustrate the approach.