Improving state estimation accuracy in active distribution networks by coordinating real‐time and pseudo‐measurements considering load uncertainty

Abstract

Measurements at the level of active distribution networks can be classified into three major groups. Data collection and transmission involve the time interval of several seconds for the first category, several minutes to 24 h for the second category, and finally, several weeks to several months for the third category. Considering this three-level classification of the measurements, state estimation by these data should deal with three kinds of heterogeneous data, but it has undesirable effects on the state estimation process, such as lower accuracy of the results. Despite the existence of heterogeneous data and different time rates, this study presents a state estimation model for updating measurement devices in each of the categories, aiming at establishing coordination among the measured data and increasing the accuracy of the estimation. The opportunities and limitations of the proposed model are compared to those of the conventional models. To illustrate the effectiveness of the proposed method, two IEEE standard systems 34-bus and 123-bus, have been used. The experimental network includes various load patterns, as well as distributed generation sources whose uncertainty is considered in the proposed model. 1Various measurement devices are currently installed on distribution networks. They are compared in Table 1.

State Estimation in Unbalanced Distribution Networks by Symmetrical Components

“…In (13a),  W refers to the solution of (12). Constraints (13b) -(13c) force the direction matrix to be in the convex hull of all rank 21 N  orthogonal matrices and by optimization, we look for extreme points of the set.…”

“…Distribution system SE can be divided into two categories of dynamic SE [10,11] and static SE [12]. In terms of static SE, the earlier research papers on distribution system SE focused on the conventional Newton-Raphson method [13,14,15].…”

Pseudo‐measurement‐based state estimation for railway power supply systems with renewable energy resources

“…In data-driven methods, [9][10][11] use artificial neural networks (ANNs) to obtain high-precision node power injection pseudomeasurement values and improve state estimation accuracy. Reference [12] considers the uncertainty of the load and improves the accuracy of state estimation through a pseudo-measurement model. References [13,14] use deep belief networks to model pseudo-measurement modelling for increasing the measurement redundancy of the system.…”

Linear Distribution System State Estimation by Hybrid Synchronized/Unsynchronized Measurements

“…DSSE is associated with a greater number of challenges compared to transmission SE due to factors such as insufficient observability, unbalanced radial configuration, and a high ratio of r/x [7,8]. Extensive work has been carried out to alleviate these challenges [9][10][11][12][13]. The observability problem is considered as a major concern due to limited metering instruments compared to the large size of distribution systems.…”

“…However, the integration of smart meters, advanced metering infrastructures (AMIs), and micro phasor measurement units (µPMUs) hold great promise for implementing DSSE [8]. The usage of pseudomeasurements and virtual measurements is primarily aimed at increasing measurement redundancy in distribution systems [9][10][11]. The unbalanced operation caused by singleor double-phase loads and untransposed lines is deemed to be another major issue.…”