Showing 1–2 of 2 results for author: Tate, J E

Line Outage Identification Based on AC Power Flow and Synchronized Measurements

Authors: Zhen Dai, Joseph Euzebe Tate

Abstract: This paper proposes a method of identifying single line outages in power systems based on phasor measurement unit (PMU) measurements and ac power flow models. In addition to the main identification algorithm, a rejection filter is introduced so that the preliminary identified results can be further processed and categorized into three types: correctly identified, misidentified and inconclusive (in… ▽ More This paper proposes a method of identifying single line outages in power systems based on phasor measurement unit (PMU) measurements and ac power flow models. In addition to the main identification algorithm, a rejection filter is introduced so that the preliminary identified results can be further processed and categorized into three types: correctly identified, misidentified and inconclusive (including correct-filtered and misidentified-filtered). The methods are systematically tested using test systems of various sizes for various PMU placements, and the results show that the proposed identification algorithm has a high identification accuracy and the proposed rejection filter is able to reduce the misidentified rate without significantly increasing the number of inconclusive cases. △ Less

Submitted 18 November, 2020; originally announced November 2020.

Comments: 5 pages, 5 figures, accepted by 2020 IEEE PES General Meeting

Hot-Starting the Ac Power Flow with Convolutional Neural Networks

Authors: Liangjie Chen, Joseph Euzebe Tate

Abstract: Obtaining good initial conditions to solve the Newton-Raphson (NR) based ac power flow (ACPF) problem can be a very difficult task. In this paper, we propose a framework to obtain the initial bus voltage magnitude and phase values that decrease the solution iterations and time for the NR based ACPF model, using the dc power flow (DCPF) results and one dimensional convolutional neural networks (1D… ▽ More Obtaining good initial conditions to solve the Newton-Raphson (NR) based ac power flow (ACPF) problem can be a very difficult task. In this paper, we propose a framework to obtain the initial bus voltage magnitude and phase values that decrease the solution iterations and time for the NR based ACPF model, using the dc power flow (DCPF) results and one dimensional convolutional neural networks (1D CNNs). We generate the dataset used to train the 1D CNNs by sampling from a distribution of load demands, and by computing the DCPF and ACPF results for each sample. Experiments on the IEEE 118-bus and \textsc{Pegase} 2869-bus study systems show that we can achieve 33.56\% and 30.06\% reduction in solution time, and 66.47% and 49.52% reduction in solution iterations per case, respectively. We include the 1D CNN architectures and the hyperparameters used, which can be expanded on by the future studies on this topic. △ Less

Submitted 20 April, 2020; originally announced April 2020.