- Title
- New state estimation techniques for smart power networks with communication constraints
- Creator
- Tai, Xin
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2013
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- Electric power networks are undergoing profound changes in recent years and receiving increasing attention from researchers in different fields. The objective is to develop a smart grid solution in the energy management system (EMS), which can enhance the efficiency, reliability, economics, and sustainability of the production and distribution of electricity in an automated fashion. With the consideration of realizing the envisioned functionalities of smart grid, massive amounts of real-time measurement data collected via a myriad of smart meters installed in different levels of the future grid are required. This will cause a huge computation and communication burden. Hence, in the development of smart grid, novel technologies should be studied to deal with this problem. To this end, two problems of state estimation in power systems are mainly considered in this dissertation: 1. The impact of communication constraints on the state estimation performance. 2. Distributed state estimation with communication constraints. Firstly, this dissertation starts with the traditional weighted least squares (WLS) estimation method which has been widely utilized in practice for two decades. However, the issue of random communication packet loss fails the system topological observability which is a necessary condition of the WLS estimation method. Hence, the maximum a posteriori estimation (MAP) method and Kalman filtering based dynamic estimation method which can overcome this numerical problem are utilized in this dissertation to address the state estimation in power system with communication constraints. Moreover, both kinds of estimation methods utilize the previous statistic information to motivate the estimation process in current time instant, which offers more accurate estimation. The expected value and the asymptotic expected value of the estimation error covariance are adopted to evaluate the performance of MAP estimation and dynamic estimation, respectively. A sequence of upper and lower bounds is proposed to approximate the asymptotic expected value. Numerical experiments are carried out using the IEEE 14-bus test system with various random communication packet loss rates, which provides a novel analysis method for engineers in practical applications. Secondly, the application of phasor measurement unit (PMU) devices in the state estimation field is studied in this dissertation. PMUs with the advantages of high sampling rate, synchronized time stamping, direct measurement of phasor and good accuracy provide a good support to real-time monitoring. Two methods of combining the traditional measurements collected by Supervisory Control and Data Acquisition (SCADA) systems and the phasor measurements obtained via PMUs in static estimation process are reviewed. And a hybrid dynamic estimator is proposed, which is capable of utilizing the high sampling rate of phasor measurements and the good redundancy of the traditional measurements. The impact of phasor measurements on the state estimation performance is analyzed under communication constraints. Based on the analysis results, an optimal PMU placement algorithm is proposed with the criteria of topological observability and estimation performance, which is able to offer a unique optimal placement solution for the power systems with various packet loss rates. Finally, distributed estimation algorithms are proposed in the last two chapters to decentralize the traditional state estimation method. The proposed distributed estimation method can offer the same global optimal performance of the traditional centralized estimation method in a finite number of iterative steps with a very low requirement of computational and communication loads. In addition, the simulation results based on IEEE standard test systems show a good robustness of the proposed distributed estimation method to communication deficiencies and subsystem asynchronism.
- Subject
- state estimation; power networks; smart grid; communication constraint; optimal PMU placement; distributed estimation
- Identifier
- http://hdl.handle.net/1959.13/940999
- Identifier
- uon:13153
- Rights
- Copyright 2013 Xin Tai
- Language
- eng
- Full Text
- Hits: 580
- Visitors: 1608
- Downloads: 373
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | ATTACHMENT01 | Abstract | 210 KB | Adobe Acrobat PDF | View Details Download | ||
| View Details Download | ATTACHMENT02 | Thesis | 3 MB | Adobe Acrobat PDF | View Details Download |