The main objective of this paper is to reintroduce some physical knowledge into the mathematical representation
of the external part while keeping its size controlled. Hence, we propose to develop a dynamic equivalent based on the
concept of scale-invariance with the aim of coping with the complexity of smart grids. As many complex systems,
smart grids are expected to be scale invariant because of the self-similarities of their hierarchic structure [27,28]. Scale
invariance means that some patterns are repeated in the hierarchical arrangement of the different levels of the smart
grid. Scale invariance shows itself by power-law relations which appear between different levels of observation of the
system. A crucial parameter to characterize the scale invariance is the power-law exponent [3]. Power-law functions
have already been used for power systems in [4] and for ICT infrastructure in [7]. Usually, power-law functions
are used for characterization purposes and the value of their exponent helps to sort out the topological properties of
networks. In the present paper, scale invariance will be introduced into the dynamic equivalent to control its size while
representing the dynamic couplings inside the smart grid.