Back to results list
Please use this identifier to cite or link to this item:
|Title:||Synchrophasor technology and its application in power system||Authors:||Wu, Ting||Advisors:||Chung, C. Y. (EE)||Keywords:||Electric power systems -- Mathematical models.
Electric power systems -- Reliability.
Electric power system stability.
|Issue Date:||2017||Publisher:||The Hong Kong Polytechnic University||Abstract:||The recent system blackouts triggered tremendous interest for wide-area situational awareness (WASA) which is based on synchrophasor measurements obtained from Phasor Measurement Units (PMUs) or PMU-enabled Intelligent Electronic Devices (IEDs). Synchrophasor technology has become an essential component in the power system for enhancing grid reliability, detecting islanded sections, aiding in system restoration, and enabling state estimator algorithms to converge with higher accuracy. These applications require accurate synchrophasor measurements from multiple synchronized devices to provide a valid picture of the dynamic behavior of the system. However, the existing commercial PMUs have poor response under dynamic conditions, such as frequency deviation, amplitude/phase angle abrupt change and modulation. This thesis is devoted to developing an advanced synchrophasor measurement algorithm which can provide accurate phasor estimation under both steady and dynamic conditions, and apply the synchrophasor technology in power systems for fault location and state estimation (SE) in order to promote the existing WASA beyond its standard purposes of monitoring, observation and post-mortem analyses. First, an intelligent phasor estimation algorithm based on random forest (RF) technology is proposed to improve estimation accuracy with reduced latency. This algorithm pre-analyses the input signals to identify step change using Lipschitz Exponent (LE) and Wavelet Transform (WT), then the negative impact (deteriorating the accuracy of phasor estimation) of which is eliminated using an adaptive data window. Next, parametric signals are categorized into three types, i.e., signals containing harmonics, inter-harmonics, and others, by a well-trained RF classification model. For the first two categories, RF regression models are developed to eliminate both harmonics and inter-harmonics in phasor measurement. Besides, Clark transform with variable phase compensation (VPC) method is proposed to precisely track the true phasor for the third type of signals.
Secondly, a fault-location technique, aimed at accelerating restoration, reducing crew repair costs, and enhancing the reliability of delivery in power systems, is developed for multi-terminal multi-section non-homogeneous transmission lines which combine overhead lines with underground power cables, by using voltage and current synchrophasors obtained from PMUs. Specifically, a faulty branch selector is first proposed to narrow down the suspected faulty area. Then, the faulty section and the exact fault location can be identified by calculating the normalized fault distance for each section on the selected faulty branch. The computational burden of the proposed analytical scheme is very low because it avoids iterative computations. Lastly, a fast state estimator and corresponding bad data (BD) processing architecture are presented to incorporate a limited number of PMUs into power system SE which plays a key role in the effective operation of power markets and enables real-time security assessment of power systems. The conventional measurements are processed by a three-stage SE method composed of two weight least squares (WLS) linear problems and a nonlinear explicit transformation in between. Then, the power system is decomposed into PMU observable and unobservable areas. The PMU observable states are calculated by a linear estimator, whereas the pseudo state values with large variances are assigned to PMU unobservable states. The results obtained from the above two estimators are then combined using the estimation fusion theory. Bad phasor measurements and bad conventional measurements in the PMU-observable area are identified and processed all at once, which can dramatically reduce the implementation time.
|Description:||PolyU Library Call No.: [THS] LG51 .H577P EE 2017 Wu
xv, 158 pages :color illustrations
|URI:||http://hdl.handle.net/10397/65266||Rights:||All rights reserved.|
|Appears in Collections:||Thesis|
Show full item record
Files in This Item:
|b29527971_link.htm||For PolyU Users||208 B||HTML||View/Open|
|b29527971_ira.pdf||For All Users (Non-printable)||2.89 MB||Adobe PDF||View/Open|
Citations as of Sep 17, 2018
Citations as of Sep 17, 2018
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.