Online Non-Iterative Estimation of Transmission Line and Transformer Parameters by SCADA Data

Abstract

Utilization of abundant measurements provided by supervisory control and data acquisition (SCADA) system has attracted increasing attention. Real-time estimation of transmission line parameters, utilizing voltage and power flow measurements provided by remote terminal units (RTUs) located at two substations across the line, has been investigated, recently. This paper improves this approach by introducing a novel exact linear reformulation of the problem, which can be solved in closed form. The distributed-parameter model of long transmission lines is considered and its parameters are estimated in a noniterative manner using RTU measurements. The method is also extended to estimate transformer series impedance and tap position by SCADA measurements, linearly. As such, the disadvantages associated with the previous iterative approach, e.g. concern over convergence, for transmission line parameters are avoided. Moreover, the novel technique for estimating transformer parameters allows to determine the tap position as well as updated transformer series impedance. Furthermore, a thorough analysis is presented to take the measurement accuracy into account. Simulation results for different transmission lines and transformers in the IEEE 118-bus test system are reported, where the result indicate successful performance of the proposed algorithms.

Nonlinear Analysis on Transmission Line Parameters Estimation From Noisy Phasorial Measurements

“…4 2500 [20] yes no ∆ρ = 0, 5% ∆ ϕ = 9 • 10 −3 rad SCADA 3.88 1000 [15] yes no Observing the Fig. 5, it is possible to conclude that the method has a stable performance even when the load is abruptly varied.…”

“…Indeed, if the number of samples is the same for the methods, a noniterative one presents a lower computational cost. However, the noniterative estimation methods present in the literature [17], [20], [22] use a large number of samples to obtain a satisfactory performance for the parameters, such methods employed more than 1000 samples in the estimation process. The presented method shows good performance even for less than 200 samples.…”

“…In a summarized description, some terms are considered Gaussian and unbiased, which is definitely not true. Some accurate statistical models for noise analysis were also proposed in the technical literature [17], [20]- [22]. For example, in reference [17], a model is developed based on realistic specifications of ITs and PMUs.…”

“…Another important issue is that this work analyzes a transmittance matrix model for a shortlength transmission line. A similar method was previously proposed in the technical literature, with an exact linear reformulation of the problem, which can be solved in closed form [20]. In such case, the distributed-parameter model for long transmission lines is considered, and its parameters are estimated in a non-iterative manner.…”

A novel approach for transient stability using interval arithmetic and optimization methods

“…In this general model, it is possible to consider the whole set of parameters as uncertain or just a subset of them (since any scalar parameter p can be represented as an interval p = [ p , p ]). Usually, in the transmission system, the line parameters are the most uncertain 26 due to temperature variation 27 or inaccurate specifications. With the increasing use of smart grids, local generation and even electric cars, loads will also contain uncertainties; the reason for this is: loads can represent a system with distributed generation and since this type of generation has low inertia or is controlled by power electronics—which have a faster dynamic than the transient stability analysis—they can be modeled as a static generation with an uncertainty.…”