A pruning tool for the multi-objective optimization of autonomous electrification systems

Despite global progresses worldwide, many people living in rural areas still have no electricity. Autonomous generation systems based on renewable energy and microgrid distribution represent a viable alternative, but their design is complex. In a recent work, the design of electrification systems is tackled through a multi-objective optimization (MO) approach that simultaneously minimizes the system cost and maximizes the energy and power supplied to consumers. However, hundreds of non-dominated solutions may result from this MO process, hindering the subsequent task of decision-makers confused by too many alternative configurations. In this framework, this work proposes a computational tool based on e-dominance and cluster analysis to prune the set of non-dominated solutions to a manageable number. The numerical experiments highlight that, regardless the size of the community to be electrified, the proposed tool successfully determines a reduced number of diverse trade-off alternatives, among which decisionmakers are able to comprehensively select their preferred option.