Managing Power Flows in SmartGrids with physically-inspired reactive agents

Managing electrical energy is nowadays a challenge of paramount importance in many countries. One of the numerous problems of this challenge is the one that consists in determining (and managing) the power flows between consumers and producers in a micro-grid (i.e. a local electrical connected network nearly isolated from the main, national level, electricity network), so as to take advantage of the renewable sources, typically solar panel and wind generator, and solicit the main grid (i.e. the global network) the least possible in order to fulfill the demand, for instance. To manage the power flows, we propose in this paper an approach based on agents that represent consumers and producers. They are moved by attractive and repulsive forces, inspired by Newtonian Physics, whose intensities depend on the amount of electrical power available by the ones and required by the others. Experimental results obtained from simulations show that this approach can manage power flows in an open system by avoiding black-out. Moreover, the results obtained show adaptability skills (i.e. producers can be added and removed in runtime).