Show simple item record

dc.contributor.advisorMoses, Paul
dc.contributor.authorFei, Wanghao
dc.date.accessioned2021-05-19T21:26:22Z
dc.date.available2021-05-19T21:26:22Z
dc.date.issued2021-05-14
dc.identifier.urihttps://hdl.handle.net/11244/329625
dc.description.abstractTo achieve the net-zero carbon dioxide emissions goal, the penetration rate of distributed energy resources has been increasing in modern power distribution networks for the past decade. Although these energy resources are environmentally friendly, they raise challenges for distribution network protection. Distribution networks are typically designed based on the single power flow direction principle, where power is flowing from the substation transformer to the load following a tree-like topology. In this way, power lines that are closer to the substation may have higher current flows. For distribution network feeder protection, particularly overcurrent relays, coordination is achieved based on the above principle, such that the downstream power lines closer to the fault have equal or higher level of fault currents compared to the upstream power lines. With distributed energy resources, the distribution network can work in either grid connected mode as is most common or islanded mode as in emerging microgrids. This indicates that the electrical topology of the distribution network can be changed in real time. Moreover, when a fault happens, the downstream relay can see higher level of fault current compared to the upstream relay, causing malfunctions of the relay, such as blinding or sympathetic tripping. The main focus of this thesis is on the development and the implementation of a new current tracing decomposition method to address the above issues. Specifically, a very detailed grid model is proposed, which has sufficient information of the current flows both from each distributed energy source to the power lines and between each distributed energy source and loads. With the results of the current flow information from the current tracing method, this research highlights the implementation of machine learning for fault current identification. Specifically, the current tracing method is taken as the kernel function that can be used to improve the performance of the support vector machine for the detection of low level faults that may be below the sensitivity of conventional overcurrent relays in the presence of DERs. This research also highlights the implementation of the new current tracing method on primary and backup protection schemes in distribution feeders. Specifically, decomposed currents are used as a substitute of the measurement currents to better coordinate the upstream and downstream relays. To demonstrate the effectiveness of the proposed method, the current tracing method is implemented in a Matlab-Simulink platform and imported to EMTP-MATLAB simulation interface. The simulation results show that using the decomposed current can improve the sensitivity and dependability of primary and backup protection in the presence of multiple DERs. It can also address the issue of protection relay blinding caused by the injection of distributed energy resources.en_US
dc.languageen_USen_US
dc.subjectPower Systemen_US
dc.subjectProtection Schemeen_US
dc.subjectCurrent Tracingen_US
dc.subjectDistributed Energy Resourcesen_US
dc.titleReal time distribution network protection considering the impact of renewable energyen_US
dc.contributor.committeeMemberJiang, John
dc.contributor.committeeMemberTang, Choon Yik
dc.contributor.committeeMemberDavis, Chad
dc.contributor.committeeMemberCai, Jie
dc.date.manuscript2021-05-12
dc.thesis.degreePh.D.en_US
ou.groupGallogly College of Engineering::School of Electrical and Computer Engineeringen_US
shareok.orcidhttps://orcid.org/0000-0002-1678-4095en_US
shareok.nativefileaccessrestricteden_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record