Figure - uploaded by Shreyasi Som
Content may be subject to copyright.
Source publication
A microgrid may be subjected to various unexpected events, such as sudden tripping of a generator/load, line outages due to faults, sudden switching of large capacitor banks, etc. Detection and classification of events play an essential role in the reliable operation, control, and restoration of microgrids. Due to low observability in microgrids, a...
Similar publications
In this paper, a low-profile, compact size, inexpensive, and easily integrable frequency reconfigurable antenna system is proposed. The proposed antenna consists of an inverted-F shape antenna, capacitors, and switching PIN diodes. The designed antenna element is fabricated on easy available and less expensive FR-4 substrate ( ε r = 4.4, tan δ = 0....
Citations
... The relevant data about microgrid is required in these approaches for the examination of relation between input and output variables to identify an event. This required data is collected through PMU measurement points and it has to be processed through machine learning or signal processing approaches for the situational awareness of the events in the microgrids (Casagrande et al., 2014;Som et al., 2022). Moreover, extraction and analysis of time-frequency characteristics has been achieved by pre-processing the input signal with the adoption of DSP approaches viz. ...
Effective microgrid control for system recovery and restoring normal operation necessitates fast event detection and implementation of remedial action (if need arises). However, fast and reliable event detection in microgrids is challenging because of low observability and inconsistencies in measurements. A novel technique is proposed in the present work for the real-time event detection and to identify the various emerging abnormalities in the microgrid. The continuous energy signature using TKEO (Teager-Kaiser Energy Operator) of the continuous varying voltage and frequency signal are extracted through μPMU. REII (Robust Event Identification Index) is constructed from these energy signatures and based on its abrupt post-event deviation from the nominal values an event is flagged in the proposed method. The proposed method is data–driven and only depends on the real-time inputs through μPMUs thus it automatically adapts the uncertainties associated with the intermittent sources of energy in the microgrid under different operating conditions. The traditional event detection techniques fail in identification of abnormalities for a microgrid connected to the transmission systems and equipped with multiple DERs such as PVDG, WG etc. To address this challenge, an integrated microgrid with multiple DERs viz. PVDG, WG and a SG (Synchronous Generator) is first developed in this work. The complexity of simultaneous operation of a static generator i.e. PVDG along with a rotor-based generator such as WG and SG is handled by the modeling the dynamic controllers of PVDG and WG for their frequency and voltage control. The simulation results depict the efficiency, accuracy and robustness of the proposed technique in terms of estimation time, event accuracy and applicability in all types of events. Moreover, the presented methodology is also compared with the four AI/ML based methods to highlight the superiority of the method.
... Due to differences in the timing of data collection by load terminals for low-voltage distribution networks, real-time measurement data is considered quasi real-time measurement data, and there are also certain differences in the collection time of this measurement data, which will affect the real-time performance of the measurement data to a certain extent [5][6]. For real-time measurement data, we use the difference algorithm to process the measurement data collected by each measurement terminal, so that they are all collected at the same time to form real-time data. ...
This project takes high-penetration distributed photovoltaic access as the background and studies the control method for low voltage and high penetration distributed photovoltaic access. Firstly, by analyzing the measured data of the low-voltage distribution network, the daily load and measured values of the low-voltage distribution network were obtained. On this basis, the topology structure of a high-penetration distributed photovoltaic grid-connected system is studied, and the voltage, current, power, and other parameters of the grid-connected system are obtained. A low-voltage and high-penetration distributed photovoltaic grid-connected control method based on a genetic algorithm is adopted. Experiments have shown that the maximum allowable capacity of this scheme is smaller than the standard value and can effectively suppress voltage fluctuations. The experimental results show that the system has good grid connection performance, which is of great significance for improving the operational stability and efficiency of low-voltage distribution systems.
... Som et al. [29] employed the event detection and bad measurement replacement model to detect multiple events in the microgrid-connected system. That employed model only uses voltage phasor and interdependency characteristics among the states to detect various events. ...
... Data-driven approach [25] has used 2 window sizes with a computation time of 132.88 s; meanwhile, PMUNET [26] has considered higher accuracy at 21 ms with a window size of 1. The recursive least square regression model [29] has reduced computation burden timing of 0.48 s. The numerical comparison between those values is presented in Table 10. ...
The power grid operators collect the data on a smart grid via the phasor measurement unit (PMU), which is placed at several points on the bus. However, the data at the PMU is prone to bad data; to classify that data, a detection scheme must be executed in the server system. The events in PMU data were subjected to several transmission conditions in the distribution system, and detecting those conditions is the major need for research. Hence to cope with that, a multi-level classification of synchrophasor’s data is proposed in this paper. An agglomerative clustering-based extreme gradient boosting (AXGBoost) classifier is proposed for a multi-level classification. The puzzle optimization method updates the probability of a data line point to the agglomerative cluster detector. The model is analyzed on IEEE 14 and 39 bus system PMU data for classification and evaluation of a model. As a result, the classification accuracy of 99.49% is obtained for IEEE 14 bus, while the accuracy for IEEE 39 bus is 99.74%, which is comparably higher than the 14 bus system. The obtained result on both systems is compared with state-of-the-art techniques such as isolation forest classifier, weighting and imbalance bagging and PMUNET.
... Thus, deviations in output generation of BESS units can provide the information about additional frequency support needed by the IRM after system disturbances. However, accurate event detection algorithms are required to detect and replace any bad data in the set of measurements to avoid miscalculation of BESS generation change [30]. Thus, an event detection and bad measurement replacement (EDBMR) model is adopted from our previous work, [30], for accurate detection of any large load/generation fluctuations in the system. ...
... However, accurate event detection algorithms are required to detect and replace any bad data in the set of measurements to avoid miscalculation of BESS generation change [30]. Thus, an event detection and bad measurement replacement (EDBMR) model is adopted from our previous work, [30], for accurate detection of any large load/generation fluctuations in the system. The details of the model can be found in [30] and are not repeated here. ...
... Thus, an event detection and bad measurement replacement (EDBMR) model is adopted from our previous work, [30], for accurate detection of any large load/generation fluctuations in the system. The details of the model can be found in [30] and are not repeated here. The EDBMR model will raise the event flag, E F , to 1, only if an event is detected in its control area; otherwise it will remain zero. ...
Due to limited reserves and 100% inverter-based resources in an islanded residential microgrid (IRM), largefrequency oscillations may arise during load/generation fluctuations. As an independent grid-forming unit, a battery energy storage system (BESS) can participate in load-frequency control (LFC) to achieve environment-friendly and reliable supply in IRMs. The formation of an interconnected islanded residential microgrid (IIRM) system and coordinated active power interchange among IRMs can help overcome BESS units limited reserve constraints. In this work, a reserve control (RC) level is proposed to be added in the conventional LFC architecture to provide frequency support to disturbance-affected areas with the help of a set of BESS units available reserve capacity. Following a significant disturbance detected, the RC level is activated based on their pre-defined contracted demands and net generation change information. The RC level also considers real-time monitoring of each microgrid's active power reserve to avoid frequency instability during emergency conditions. System performance with the proposed LFC model is analyzed with extensive tests conducted on a CIGRE TF C6:04:02 benchmark IIRM system. The proposed algorithms are also validated using a real-time digital simulator with hardware-in-the-loop setup.
... The method can distinguish between internal, external, and severe no-fault circumstances. [194] This method is dependable and flexible for different microgrid networks without requiring system model data or depending on particular disruption characteristics because it uses a regressive vector model, an indicator function to separate events from inaccurate measurements. ...
The emerging smart-grid and microgrid concept implementation into the conventional power system brings complexity due to the incorporation of various renewable energy sources and non-linear inverter-based devices. The occurrence of frequent power outages may have a significant negative impact on a nation’s economic, societal, and fiscal standing. As a result, it is essential to employ sophisticated monitoring and measuring technology. Implementing phasor measurement units (PMUs) in modern power systems brings about substantial improvement and beneficial solutions, mainly to protection issues and challenges. PMU-assisted state estimation, phase angle monitoring, power oscillation monitoring, voltage stability monitoring, fault detection, and cyberattack identification are a few prominent applications. Although substantial research has been carried out on the aspects of PMU applications to power system protection, it can be evolved from its current infancy stage and become an open domain of research to achieve further improvements and novel approaches. The three principal objectives are emphasized in this review. The first objective is to present all the methods on the synchro-phasor-based PMU application to estimate the power system states and dynamic phenomena in frequent time intervals to observe centrally, which helps to make appropriate decisions for better protection. The second is to discuss and analyze the post-disturbance scenarios adopted through better protection schemes based on accurate and synchronized measurements through GPS synchronization. Thirdly, this review summarizes current research on PMU applications for power system protection, showcasing innovative breakthroughs, addressing existing challenges, and highlighting areas for future research to enhance system resilience against catastrophic events.
... Similarly, a data-driven method to differentiate measurement anomalies from events in microgrids using PMUs is proposed in [114]. The changes are detected by comparing the changes in voltage phasor measurements of a PMU with other PMUs. ...
Increasing power demand, aging distribution systems and concerns towards greenhouse gas emissions have resulted in the increased occurrence of distributed generation (DG) within distribution networks. The conventional protection methods designed for passive radial distribution networks may become redundant with large bidirectional power flow and dynamic network topology. Additionally, the anti-islanding protection currently employed limits the benefits of wide-scale DG installation and autonomous operation. The microgrid concept can solve these problems, but several challenges must be overcome before practical implementation. Besides bi-directional power flow, the vast variance between the fault current in grid-connected and autonomous mode and the arbitrary output impedance of the inverter-interfaced DG units in fault conditions and current limiting mode pose a challenge to the protection schemes that use traditional overcurrent protection devices. Many researchers have proposed various techniques, but a robust protection scheme capable of protecting microgrids against different faults for both modes of operation under dynamic network topologies and being financially viable is still to be developed. Hence, the main objective of this paper is to critically review various AC microgrid protection methods proposed in the literature, focusing on analysing the recent protection approaches using modern intelligent techniques. Open research problems and future research trends in AC microgrid protection are also presented in this research.
... Event detection has been explored significantly in the electrical distribution grid. Existing literature has presented data-driven approaches based on statistical [2], [7], [8], physics-derived [9], [10], machine learning-based [11], [12], [13], [14], [15], [16], [17], and graph-based [18], [19], [20] methods. Statistical methods involve the concept of absolute deviation around the median with consideration of dynamic window sizes [2]. ...
The expansion in technology and attainability of a large number of sensors has led to a large amount of real-time streaming data. The real-time data in the electrical distribution system is collected through distribution-level phasor measurement units referred to as micro-PMU (μPMU) which report high-resolution phasor measurements comprising various event signatures that provide situational awareness and enable a level of visibility into the distribution system. These events are infrequent, unscheduled, and uncertain; it is a challenge to scrutinize, detect, and predict the occurrence of such events. In this paper, we seek to address these problems by developing an unsupervised physics dynamics-based approach to detect anomalies in the μPMU data, then using an unsupervised clustering algorithm to categorize events based on their pattern and simultaneously predict the events using system governing equations. We propose a data-driven approach based on the Hankel alternative view of the Koopman (HAVOK) analysis and named it DynamoPMU, to analyze the underlying dynamics of the distribution system by representing them in a linear intrinsic space. The key technical idea is that the proposed method separates out the linear dynamical behavior pattern and intermittent forcing (anomalous events). Further, we cluster the detected events using the Gaussian Mixture Model (GMM) and predict the measurement data and anomalous events using the Sparse Identification of Nonlinear Dynamic (SINDy) model. We demonstrate the efficacy of our proposed framework through analysis of real μPMU data taken from the Lawrence Berkeley National Laboratory (LBNL) distribution grid.
This work presents a new algorithm for detecting and classifying data anomalies in operational measurements using statistical, clustering, and outlier-based approaches. Base detectors explored in this work includes density-based spatial clustering of applications with noise, K-Means, local outlier factor, feature bagging, and robust random cut forests using real distribution system datasets. An ensemble approach is proposed to achieve high detection accuracy and precision compared with any of the base detector and with less dependency on hyperparameter tuning. Also, developed ensemble architecture can integrate additional base detectors. In addition, a simplistic anomaly classification approach is developed, utilizing the clustering concept, while considering the physics of the power distribution systems. The developed schemes are rigorously tested and validated using data from multiple distribution phasor measurement unit devices in the Bronzeville community microgrid, with a diverse set of events and distributed energy resources at dispersed locations. Performance analysis using three test cases are provided to showcase superiority of the proposed approaches.
Secure state estimation is becoming more popular due to the inherent vulnerabilities of communication networks in essence, which could give rise to potential data leakage and manipulation of microgrids. The paper addresses the issue of secure distributed state estimation for a class of microgrids with potential outliers occurring in sensor measurements. First, a secure distributed estimator is constructed by introducing both an artificial saturation rule to achieve outlier resilience and a variable decomposition strategy to safeguard data security, where the generated dynamic key is a time-varying sequence satisfying the predetermined constraint. Deep variance analysis is carried out to profoundly disclose the relationship between private and public estimation error covariance, in accordance with the employed decomposition rule. An upper bound of error covariance is determined by two sets of recursive matrix equations in contrast to that of traditional distributed estimation. Furthermore, the desired estimator gains are obtained recursively with the aid of optimizing the upper bound obtained above. In the end, a simulation example is proposed to confirm the effectiveness and security of the proposed algorithm.