Adaptive Facet Selection in Multidimensional Hosting Capacity Region Assessment

The hosting capacity region defines a joint feasible region in the operational space involving various installations. By facilitating their coordinated operation, this region succeeds in tapping into the grid's power delivery potential. Meanwhile, when approximating this region using an expansive polytope, facet selection should be carefully considered. Random facet selection will cause the Matthew effect, indicating that a facet which was more frequently selected owns a higher likelihood of being selected again. Such effect eventually harms the region expansion efficiency. Beyond static facet selection measures, this paper proposes adaptive measures for further improved assessment performances. The adaptive scheme is engineered to cyclically use original measures, thereby merging and leveraging their potentials on efficient facet selection. Relevant 3-dimensional region assessment experiments are conducted on a 10.5 kV Dutch grid case, which is modelled on Pandapower toolkit. The results demonstrate that, compared to static alternative measures, adaptive measures contribute to larger region space consistently, with potential region space gain being up to 62.2%. This validates the superior efficacy of adaptive mechanism in facet selection for region assessment.