Implementing an auction market for regulation service using software agents

Software agent based systems have long been used in the design and implementation of complex systems. This paper discusses a framework for the regulation service auction market for investigating the behaviors of software agents representing the various participants including the auction coordinator (ISO). The main emphasis is to consider the profit maximizing bidding strategies employed by the power suppliers to gain from the competitive market setup. The Java Multi-Agent Development Kit (Madkit) is used for the auction simulations and Java Native Interface (JNI) methods are employed to invoke Matlab computations. Finally, two examples are shown for procurement of regulation service involving five suppliers over a twenty-four hour auction period.