Sparsity Based Approaches for Distribution Grid State Estimation - A Comparative Study

Abstract

The power distribution grid is typically unobservable due to a lack of measurements. While deploying more sensors can alleviate this issue, it also presents new challenges related to data aggregation and the underlying communication infrastructure. Therefore, developing state estimation methods that enhance situational awareness at the grid edge with compressed measurements is critical. For this purpose, a suite of sparsity-based approaches that exploit the correlation among states/measurements in spatial as well as temporal domains have been proposed recently. This paper presents a systematic comparison and evaluation of these approaches. Specifically, the performance and complexity of spatial methods (1-D compressive sensing and matrix completion) and spatio-temporal methods (2-D compressive sensing and tensor completion) are compared using the IEEE 37 and IEEE 123 bus test systems. Additionally, new robust formulations of these sparsity-based methods are derived and shown to be robust to bad data and network parameter uncertainties. Among the sparsity-based approaches, compressive sensing methods tend to outperform matrix completion and tensor completion methods in terms of error performance.

Overview of Loss Sensitivity Analysis in Modern Distribution Systems

“…This information can then be transferred to distribution system state estimation (DSSE). DSSE provides the asset management center with the required information about nodal voltages, power measurements and line current flows [64] [65] [66]. Such information can be used as the input to outage management systems (OMS) [67].…”

Recursive Gaussian Process over graphs for Integrating Multi-timescale Measurements in Low-Observable Distribution Systems

“…Tensor completion fills the missing elements in a tensor by exploiting the spatio-temporal correlation of the measurements [20]. A comparative analysis of these sparsity-based approaches along with their robust formulations was proposed in [21]. Authors in [22] use PMU and SCADA measurements for DSSE.…”

“…While not the primary focus or contribution of our work, we provide a brief summary of the matrix completion based DSSE proposed in [18] for the sake of completeness. Unlike [18], [21] that assumes time synchronized subset of measurements, here the GP based reconciled measurements are used within the Matrix completion (MC) based DSSE. MC based DSSE estimate the spatial states of the network (i.e., the voltage phasors and power injections of all the buses at a single instant of time) by exploiting the sparsity of raw measurements.…”

“…(60) and (61) captures the linearized power-flow constraint relating voltage phasors v and voltage magnitude |v| to the power measurements as given in [33]. More details about the matrix completion based DSSE can be found in [18], [21]. Additionally, the impact of uncertain topology on matrix completion based DSSE is considered in [8], [21].…”

O Uso de AutoEncoders como Fomentador de Medidas em Sistemas de Distribuição

“…Alguns outros autores preferem utilizar as medidas existentes para obter o estado diretamente. Os autores de [Dahale et al 2020] mostram como o fluxo de potência linear e programac ¸ão linear podem ser combinados para a obtenc ¸ão do estado da rede. Em outra linha, [Mestav et al 2019] expõe a utilizac ¸ão de Simulac ¸ões de Monte Carlo e modelos de redes neurais profundas para obter um melhor resultado em relac ¸ão à detecc ¸ão de erros de medic ¸ão e a estimac ¸ão de estado na totalidade.…”