Decision Tree Based Anomaly Detection for Remedial Action Scheme in Smart Grid using PMU Data

Singh, Vivek Kumar; Manimaran, G.

doi:10.1109/pesgm.2018.8586159

Cited by 38 publications

(17 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yan et al [18] compared SVM, KNN, and extended nearest neighbor (ENN) for detecting both a direct and stealth FDI attack in smart grids. Singh et al [19] proposed a decision tree-based anomaly detection approach based on PMU measurements to distinguish the normal tripping from power line faults and malicious attacks tripping the physical relays. An Adaboost-based classification model using the random forest as the base classifier using individual PMU data was developed in [20] for detecting power system disturbances and cyber attacks.…”

Section: Related Workmentioning

confidence: 99%

Detecting Cyber Attacks in Smart Grids Using Semi-Supervised Anomaly Detection and Deep Representation Learning

Rasband

Zheng

et al. 2021

Information

View full text Add to dashboard Cite

Smart grids integrate advanced information and communication technologies (ICTs) into traditional power grids for more efficient and resilient power delivery and management, but also introduce new security vulnerabilities that can be exploited by adversaries to launch cyber attacks, causing severe consequences such as massive blackout and infrastructure damages. Existing machine learning-based methods for detecting cyber attacks in smart grids are mostly based on supervised learning, which need the instances of both normal and attack events for training. In addition, supervised learning requires that the training dataset includes representative instances of various types of attack events to train a good model, which is sometimes hard if not impossible. This paper presents a new method for detecting cyber attacks in smart grids using PMU data, which is based on semi-supervised anomaly detection and deep representation learning. Semi-supervised anomaly detection only employs the instances of normal events to train detection models, making it suitable for finding unknown attack events. A number of popular semi-supervised anomaly detection algorithms were investigated in our study using publicly available power system cyber attack datasets to identify the best-performing ones. The performance comparison with popular supervised algorithms demonstrates that semi-supervised algorithms are more capable of finding attack events than supervised algorithms. Our results also show that the performance of semi-supervised anomaly detection algorithms can be further improved by augmenting with deep representation learning.

show abstract

Section: Related Workmentioning

confidence: 99%

Detecting Cyber Attacks in Smart Grids Using Semi-Supervised Anomaly Detection and Deep Representation Learning

Rasband

Zheng

et al. 2021

Information

View full text Add to dashboard Cite

show abstract

“…The cyber physical system (CPS) federated testbed works as a driving force to enable the pipeline from state-ofthe art research work through the transition to industry by experimental testing and validation. In this work, we have presented a cyber (network) based federated testbed; inspired by the NASPINet architecture, to evaluate the applied anomaly detector, as proposed in [6], in terms of network latency, communication bandwidth, and accuracy rate. We have implemented the realistic data integrity attacks targeting physical relays and substation RTUs, located at the ISU Power Cyber lab.…”

Section: Discussionmentioning

confidence: 99%

“…The attacker closes the legitimate RTU program and initiates a python scripted malicious logic routine which periodically sends control signals to the simulator targeting the generator (Gen 6) to initiate ramp and pulse attacks. We have also simulated three phase to ground faults followed by the normal tripping of the line L16-17 to simulate physical disturbances; and multiple simulations are performed for different cases as discussed in [6]. It is appropriate to note that the due to the space limitation, we are not discussing the details of different scenarios required for generating the labeled datasets.…”

Section: Federation Architecture and Detection Methodologymentioning

confidence: 99%

“…Specifically, we have focused on the SCADA-Synchrophasor integrated, control center based remedial action scheme, which receives PMU measurements including line flows, relay status and generator output to detect disturbances, and SCADA based, closed loop control signals are employed to reduce the generation, if necessary, to prevent the thermal overloading. More details are provided in [6].…”

Section: Overview and Related Workmentioning

confidence: 99%

“…In an earlier effort, we have presented the notion of decision tree based anomaly detection using differential PMU features for detecting a data integrity attack (malicious tripping) in the context of WAP security [6]. In this work, we have proposed the cyber(network) based federated testbed architecture, inspired by the NASPInet architecture, to analyze the widearea network traffic, and to evaluate the performance of the anomaly detector using a realistic platform for detecting cyberattacks.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Evaluation of Anomaly Detection for Wide-Area Protection Using Cyber Federation Testbed

Singh

Manimaran

Porschet

et al. 2019

2019 IEEE Power &Amp; Energy Society General Meeting (PESGM)

Self Cite

View full text Add to dashboard Cite

Cyber physical security research for smart grid is currently one of the nation's top R&D priorities. The existing vulnerabilities in the legacy grid infrastructure make it particularly susceptible to countless cyberattacks. There is a growing emphasis towards building interconnected, sophisticated federated testbeds to perform realistic experiments by allowing the integration of geographically-dispersed resources in the dynamic cyber-physical environment. In this paper, we present a cyber (network) based federation testbed to validate the performance of an anomaly detector in context of a Wide Area Protection (WAP) security. Specifically, we have utilized the resources available at the Iowa State University Power Cyber (ISU PCL) Laboratory to emulate the substation and local center networks; and the US Army Research Laboratory (ARL); to emulate the regional control center network. Initially, we describe a hardware-in-the loop based experimental setup for implementing data integrity attacks on an IEEE 39 bus system. We then perform network packet analysis focusing on latency and bandwidth as well as evaluate the performance of a decision tree based anomaly detector in measuring its ability to identify different attacks. Our experimental results reveal the computed wide area network latency; bandwidth requirement for minimum packet loss; and successful performance of the anomaly detector. Our studies also highlight the conceptual architecture necessary for developing the federated testbed, inspired by the NASPI network.

show abstract

Recent study advances in the Department of Polymer Science and Engineering at Zhejiang University

Dong

Gao

2022

Journal of Polymer Science

View full text Add to dashboard Cite

The polymer science at Zhejiang University (ZJU) is well known for its foremost origin of study and education of polymers in China. According to the latest US News rankings, it is ranked third in the subject of Polymer Sciences. The Department of Polymer Science and Engineering (DPSE) is one of the earliest academic institutions devoted to polymers at ZJU. Under the leadership of Prof. Zhi-Quan Shen and Prof. Jia-Chong Shen, academicians of the Academia Sinica, 54 faculty members (including 28 professors, 26 associate professors, and senior researchers) are now working in education and scientific research of polymer sciences, polymer composites and biomedical macromolecules. Our faculty is among the finest all over the nation, leading polymer science and engineering research. Our department has been developed as a collaborative community of faculties, students, and alumni dedicated to research developments and solving important technical questions with three decades of endeavors.In this special issue of ZJU, we aim to share a selection of recent research activities from our faculties, students, and alumni, classified into the following six topics: controlled polymerization, polymer physics and rheology, biomedical polymers, polymer membranes and films, organic and polymeric photo-electronic functional materials, and aggregationinduced emission (AIE) and graphene materials.The topic of controlled polymerization illustrates the fabrication of fine polymers with sophisticated polymerization techniques. A protocol of using readily available raw materials of ethylene oxide, propylene oxide, and carbonyl sulfide is developed by Zhang et al. to synthesize sulfur-rich poly(thioether)s that exhibit T m values ranging from 76 C to 144 C and have high refractive index up to 1.68. Du et al. synthesize a series of ABCBA symmetrical amphiphilic pentablock terpolymers with various block lengths two-step atom transfer radical polymerization. These polymers can form high-genus vesicles in dilute aqueous solutions. In another study, Ling et al. design the synthesis of P(EVL-alt-EDT)-g-PEG combining thiol-ene click polymerization and ring-opening grafting reaction. The obtained polymers consist of a hydrophobic backbone of alternative copolymers of disubstituted

show abstract

Decision Tree Based Anomaly Detection for Remedial Action Scheme in Smart Grid using PMU Data

Cited by 38 publications

References 8 publications

Detecting Cyber Attacks in Smart Grids Using Semi-Supervised Anomaly Detection and Deep Representation Learning

Detecting Cyber Attacks in Smart Grids Using Semi-Supervised Anomaly Detection and Deep Representation Learning

Evaluation of Anomaly Detection for Wide-Area Protection Using Cyber Federation Testbed

Recent study advances in the Department of Polymer Science and Engineering at Zhejiang University

Contact Info

Product

Resources

About