Intrusion detection system using Online Sequence Extreme Learning Machine (OS-ELM) in advanced metering infrastructure of smart grid

Abstract

Advanced Metering Infrastructure (AMI) realizes a two-way communication of electricity data through by interconnecting with a computer network as the core component of the smart grid. Meanwhile, it brings many new security threats and the traditional intrusion detection method can’t satisfy the security requirements of AMI. In this paper, an intrusion detection system based on Online Sequence Extreme Learning Machine (OS-ELM) is established, which is used to detecting the attack in AMI and carrying out the comparative analysis with other algorithms. Simulation results show that, compared with other intrusion detection methods, intrusion detection method based on OS-ELM is more superior in detection speed and accuracy.

Cyber Security in Smart Grid

“…Perhaps information on energy bills, trends, registration, labelling, geographic location of historical reports, consumer information and emails. The operational data can be real-time current and voltage values, transformer tap-changers, and capacitor banks, current loads from the transformer power supplies, fault positions, relay status, and switch status [8]. Operational data requires a high level of security to protect smart grid systems from any vulnerabilities and attacks that could cause a power outage.…”

“…A smart grid uses innovative products and services, along with smart technologies for monitoring, control, communication and self-regeneration [10]. A Smart Grid is an electric grid that can efficiently integrate the behaviour and actions of all users connected to it -generators, consumers and those who do both -in order to guarantee economically efficient and sustainable energy systems with low losses and high levels of quality and security of supply and security [8]. Since then, network cyber security has become one of the main points on the government and industrial agenda, with the recognition that the more interconnected the network becomes, the greater the attack surface.…”

The Paradigm Revolution in the Distribution Grid: The Cutting-Edge and Enabling Technologies

“…The control actions are decided based on the operating states of the power system. A self-healing scheme based on operating state classification is proposed [38] Event detection AMI Random matrix theory [39] Anomaly detection and location AMI Spectral clustering [28] Phase identification of smart meters /customers AMI Online Sequence Extreme Learning Machine (OS-ELM) [40] Intrusion detection which is used to detect the attack in AMI AMI Wavelet transform [26] Load curve clustering algorithm AMI Multi-instance clustering algorithm [41] Consumer Segmentation AMI Euclidean distance-based anomaly detection schemes [42] Covert cyber deception assault-detection AMI Ensemble Classifier Bootstrap Aggregation [43] Intrusion detection AMI Data Mining of Code Repositories (DAMICORE) [44] Islanding detection of synchronous distributed generators Micro-PMU OPTICS [45] Segment data and finds the outliers in the segmented data Micro-PMU CNN [46], RNN [46] Power system transient disturbance classification Micro-PMU Multi-SVM [47] Data driven event classification Micro-PMU Multi-SVM and PCA [48] Disruptive event classification Micro-PMU Auto-encoder together with softmax classifier [48] Disruptive event classification Micro-PMU kernel Principle Component Analysis (kPCA) [49] Abnormal event detection Micro-PMU for distribution networks with DGs in [33]. The operating states of the system are determined based on the performance index, which is used as an indicator of system stability, reliability, and security.…”

Fuzzy Multi-Objective Optimization Algorithms for Solving Multi-Mode Automated Guided Vehicles by Considering Machine Break Time and Artificial Neural Network

“…Moreover, as long as new training samples are obtained, OS-ELM updates the network weights recursively. This network weight updating algorithm lacks flexibility to adjust parameters according to actual conditions, and at the same time, it is easy to increase unnecessary computation time [28]. In order to solve the problems existing in OS-ELM effectively, an improved OS-ELM named as adaptive OS-ELM is proposed.…”