This thesis proposes two novel algorithms to analyze whether the power system loses synchronism subsequent to credible contingencies. The two algorithms are based on the concept of Lyapunov exponents (LEs) and the Prony analysis respectively.
The concept of LEs is a theoretically sound technique to study the system stability of nonlinear dynamic systems. The LEs measure the exponential rates of divergence or convergence of trajectories in the state space. Considering the higher computational burden associated with the convergence of the true LEs, a modified algorithm is proposed to study the transient stability of the post-fault power system. It is shown that the finite-time LEs calculated by the modified algorithm accurately predicts the said stability.
If the power system is transient stable, the rotor angle trajectories of the post-fault system exponentially decay with time. The damping ratios of the dominant oscillatory modes present in these power swings provide the indication on the oscillatory stability. The improved Prony algorithm presented in the thesis can be used to identify the oscillatory stability of the power system subsequent to a contingency.
It is shown that that these new algorithms can be used in two applications in power systems, online dynamic security assessment and online oscillations monitoring. The proposed algorithm for rotor angle security assessment first uses the LEs-based algorithm to identify the transient stability. The stable cases are then processed by the improved Prony algorithm. The proposed online oscillations monitoring algorithm uses an event-detection logic and a parallel filter bank before applying the improved Prony algorithm on the measured response to extract the dominant oscillatory modes and to determine their frequencies and damping ratios.
The suitability of the two algorithms for the aforementioned applications is investigated using different case studies. It is shown that the computational burdens of the two algorithms are acceptable for the online applications. Furthermore, the oscillations monitoring algorithm, extracts only the dominant modes present in the input signal, extracts both low-frequency inter-area modes and sub-synchronous modes, and performs well under noisy conditions. These features make it more appropriate for wide-area monitoring of power system oscillations using synchronized measurements. / February 2016
| Identifer | oai:union.ndltd.org:MANITOBA/oai:mspace.lib.umanitoba.ca:1993/30972 |
| Date | 10 December 2015 |
| Creators | Wadduwage, Darshana Prasad |
| Contributors | Annakkage, Udaya (Electrical and Computer Engineering) Wu, Christine (Mechanical Engineering), Narendra, Krish (Electrical and Computer Engineering) Thulasiram, Ruppa (Computer Science) Venayagamoorthy, Ganesh Kumar (Clemson College of Engineering and Science) |
| Source Sets | University of Manitoba Canada |
| Detected Language | English |
Page generated in 0.002 seconds