An Active Distribution Network Equivalent Derived From Large-Disturbance Simulations With Uncertainty

Abstract

A reduced-order, "grey-box" model of an active distribution network, intended for dynamic simulations of the transmission system, is derived. The network hosts inverter-based generators as well as static and motor loads, whose dynamic parameters are affected by uncertainty. This issue is addressed using Monte-Carlo simulations. The parameters of the equivalent are adjusted to match as closely as possible the average of the randomized responses, while their dispersion is accounted for through the weights of the weighted least-square minimization. A procedure is used to remove from the identification the parameters with negligible impact. To avoid over-fitting, the equivalent is tuned for multiple large-disturbance simulations. A recursive procedure is used to select the smallest possible subset of disturbances involved in the least-square minimization. Simulation results with a 75-bus MV test-system are reported. They show that the equivalent is able to reproduce with good accuracy the discontinuous controls of inverter-based generators, such as reactive current injection and tripping. Index Terms-active distribution network, inverter-based generator, dynamic equivalent, grey-box model, uncertain dynamic model, Monte-Carlo simulations, LASSO method.

A Dynamic Equivalent of Active Distribution Network: Derivation, Update, Validation and Use Cases

“…This section summarizes the ADN equivalent derivation. The interested reader can find more details in to [16], [17].…”

“…Q e (θ, j, k) from µ Q (j, k)) is smaller than σ P (j, k) (resp. σ Q (j, k)) on average over discrete times k [16].…”

“…A metaheuristic, derivative-free optimization method is used to solve the minimization problem (1)-(4). Among numerous methods, an evolutionary algorithm, namely Differential Evolution (DE) [24] has been chosen, for reasons exposed in [16].…”

“…A variant of the LASSO technique has been presented in [16] whose purpose is to identify the components of an initial θ vector which are "significant". The others have negligible impact and can thus be assigned default values and be removed from θ.…”

Experimental Validation of a Dynamic Equivalent Model for Microgrids

“…In [17] a reduced order linear model is used to control a microgrid. The idea beyond the equivalent modeling approach is define and identify an Equivalent Dynamic Model (EDM), usually composed by a set of differential equations, able to reproduce the system dynamics [18]- [20]. Generally, measurement-based methods [21]- [22] are preferred to other approaches since they do not require a detailed a priori knowledge of the real system.…”