One of the most important sectors of modern society is, undoubtedly, the energy sector and
its systems of generation, transmission and distribution. These electric systems are complex
and hard to handle, requiring applications of sophisticated techniques to make energy
supply stable and minimally reliable. Smart grids is an area that has been developing
along last years, which incorporates features to electrical grids that improve the quality of
its services, such as real-time data control, digital communication between different parts
of the system, self-healing, and others. There are many researches developed that study
the self-healing problem using many techniques. The self-healing, so, is a functionality
that may bring to the electrical grids great advances in the way to provide a quality and
continuous service. In the scope of power distribution grids, this work proposes a model of
multiagent system (MAS) to deal with the self-healing problem of smart grids in power
distribution level. A MAS was developed to simulate and handle the operations, failures
and self-healing of electrical grids. The self-healing follows a four-step algorithm and takes
decisions based on a power-flow algorithm. To evaluate the proposed MAS, two models of
electrical grid were used: the first one was designed for this research, with 8 nodes, and the
second one was proposed in the work of Baran e Wu (1989), with 33 nodes. The results
were obtained by computational simulations and tabulated to analysis. These results show
that MAS proposed is able to perform the grid self-healing through the four-step algorithm
execution, emphasizing that the distributed way how the agents coordinates themselves
to perform the needed operations is a differential in relation to the response time and
computational effort required by the system.
|
