ANFIS microgrid energy management system synthesis by hyperplane clustering supported by neurofuzzy min–max classifier

A novel energy management system (EMS) synthesis procedure based on adaptive neurofuzzy inference systems (ANFISs) by hyperplane clustering is investigated in this paper. In particular, since it is known that clustering input–output samples in hyperplane space does not consider clusters’ separability in the input space, a Min–Max classifier is applied to properly cut and update those hyperplanes defined on scattered clusters in order to refine the ANFIS membership functions. Furthermore, three different clustering techniques have been compared for the ANFIS rule synthesis as well, both with and without considering the classifier support. The procedure under analysis has been applied for designing a microgrid EMS equipped with a photovoltaic generator and an energy storage system (ESS). The EMS is in charge of intelligently defining how to redistribute the prosumer energy balance between the ESS and the connected grid in order to maximize the profit generated by the energy exchange with the grid, assuming a time of use energy price policy. Results on real-world data show very interesting performances, close to optimal values evaluated with a mixed integer linear programming problem formulation by approximately 12%. Moreover, the contribution of the Min–Max classifier improves the EMS performance by approximately 50% with respect to the same algorithm without refining fuzzy rules by the classification step.