A Dynamic Mode Decomposition Framework for Global Power System Oscillation Analysis

Abstract

Abstract--A global multiscale method based on a dynamic mode decomposition (DMD) algorithm to characterize the global behavior of transient processes recorded using wide-area sensors is proposed. The method interprets global dynamic behavior in terms of both, spatial patterns or shapes and temporal patterns associated with dynamic modes containing essentially singlefrequency components, from which the mode shapes, frequencies and growth and decay rates of the modes can be extracted simultaneously. These modes are then used to detect the coherent and dominant structures within the data.The technique is well suited for fast wide-area monitoring and assessment of global instability in the context of modern data fusion-based estimation techniques. Results of the application of the proposed method to large, high-dimensional data sets are encouraging.

An interactive tool for measurement-driven modal analysis of large-scale power systems

“…The work in [3] and [4], both show that the DMD algorithm is significantly faster than methods such as Prony. In this tool, modal analysis can be performed fairly quick, and can be done repeatedly using a sliding time-window approach.…”

“…It originated from the field of Fluid Dynamics ( [1], [2]), and has recently been applied to power systems. DMD has exhibited good accuracy, and also its capability in being significantly faster than Prony ( [3]). The work in [4] has continued to build on the work in [3], and has been able to further demonstrate DMD's computation speed and scalability in accommodating more measurement signals and higher sampling rates.…”

“…DMD has exhibited good accuracy, and also its capability in being significantly faster than Prony ( [3]). The work in [4] has continued to build on the work in [3], and has been able to further demonstrate DMD's computation speed and scalability in accommodating more measurement signals and higher sampling rates. As part of ongoing research efforts, an interactive modal analysis tool has been designed.…”

A tutorial on data‐driven eigenvalue identification: Prony analysis, matrix pencil, and eigensystem realization algorithm

“…While the 2012 Taskforce report offers a great guideline on the three methods, a thorough examination on their principles, multiple data handling, noise resilient techniques, and applications in large‐scale power grid dynamics and power electronic converter penetrated system dynamics is of great importance. Further, data‐driven modeling technique of dynamic mode decomposition (DMD) and Koopman operator, which has been developed for fluid dynamics, was recently introduced to the power systems . DMD allows mode identification and coherent generator identification, and it also relies on shifted Hankel matrix construction and singular value decomposition (SVD).…”

Wide area phase angle measurements for islanding detection — An adaptive nonlinear approach

“…In this paper, multiple locations are employed in case the reference site fails or becomes islanded itself. The concept of utilizing multiple PMU locations to improve the tracking capabilities of power system dynamics has been published recently for power system inter-area oscillation monitoring [24]. To the best of our knowledge, analyzing the phase angle difference from multiple locations simultaneously, with an 'adaptive multiple cross-reference' approach, has not been studied previously in the islanding protection domain.…”