A Multi-Agent Clustering-based Approach for the Distributed Planning of Wind Generators

In this paper, a new approach for the distributed planning of the wind distributed generators (DGs) in active distribution networks and microgrids is proposed. First, an electrical distance matrix (EDM) is obtained for the distribution system. Having obtained the EDM, a maximization problem is formulated and solved for the optimal clustering of the network. In the second stage, agents are assigned to each cluster, and a multi-objective optimization problem (MOOP) is formulated for optimal planning of wind DGs and assigned to a head agent; the objective functions in the MOOP are equal to the number of agents, and the objective function of an agent is composed of annual energy losses. A voltage improvement index for each cluster is also calculated to measure the improvement in the voltage profile. IEEE 37-node test feeder and two test cases are taken into account for the study. The results show that the losses are reduced in each cluster, and the voltage profile is also enhanced at the same time.