Algorithms for Selecting and Interconnecting Switches to Automate Power Grids Considering Continuity Indexes and Reliability

In this paper, a method is presented for the deployment of switch clusters to automate distribution grids. The proposed methodology consists in choosing feeders of a self-healing system based on their performances in relation to the System Average Interruption Duration Index, the number of consumers, the compensation and to the probabilities of the current to exceed the pickup limit. To this end, it is proposed to consider statistical analysis of current values using decision theory and binary linear programming to determine the priorities of switch clusters. In addition, the problem of allocating communication links for commanding automated breakers in a distribution power grid is also addressed. The application of a multi-objective genetic algorithm is considered for optimizing link choice where costs and network reliability are objective functions. A novel heuristic is proposed for attributing reliability values to the links in terms of the subjacent power grid, inducing optimization convergence toward network topologies in which breakers at areas with higher fault indexes receive more communication resources.