Rajender Kumar Beniwal

To combat cyber threats in the smart grid, an intrusion detection system can be integrated into the advanced metering infrastructure. Anomaly-based intrusion detection can detect even the tiniest changes in the parameter under investigation, whereas signature-based intrusion detection only recognises known attacks. The growing usage of smart grids necessitates the classification, identification, and implementation of countermeasures to threats. At the absolute least, smart grids must be protected against cyberattacks; thus, the highest level of information security must be offered. As a result of digitisation and the usage of more smart applications, the research looked at a variety of attack types, smart grid assaults, and major cyber threats on the voltage regulation. Machine learning techniques that analyse data in real time and formulate patterns to recognise an attack and scan through huge data for anomalies can be implemented into the advanced metering infrastructure (AMI) for intrusion detection for anomaly-based intrusion detection. The comparative test study done for the research found that the proposed method, median absolute deviation for anomaly identification in smart metering datasets, produced the most accurate and precise differentiations with the highest accuracy and precision. The median absolute deviation (MAD) algorithm model is trained using test data, and raw predictions are made, before true data are used to derive final test result parameters, precision, recall, and F1 scores. The methodology of the entire study is discussed in this paper, as well as how the MAD algorithm is best suited for anomaly-based intrusion detection, as well as result comparisons of other machine learning algorithms.

Recently, power quality (PQ) issues have drawn considerable attention of the researchers due to the increasing awareness of the customers towards power quality. The PQ issues maintain its pre-eminence because of the significant growth encountered in the smart grid technology, distributed generation, usage of sensitive and power electronic equipments with the integration of renewable energy resources. The IoT and 5G networks technologies have a number of advantages like smart sensor interfacing, remote sensing and monitoring, data transmission at high speed. Due to this, applications of these two are highly adopted in smart grid. The prime focus of the paper is to present an exhaustive survey of detection and classification of power quality disturbances by discussing signal processing techniques and artificial intelligence tools with their respective pros and cons. Further, critical analysis of automatic recognition techniques for the concerned field is posited with the viewpoint of the types of power input signal (synthetic/real/noisy), pre-processing tools, feature selection methods, artificial intelligence techniques and modes of operation (online/offline) as per the reported articles. The present work also elaborates the future scope of the said field for the reader. This paper provides valuable guidelines to the researchers those having interest in the field of PQ analysis and exploring the better methodologies for further improvement. Comprehensive comparisons have been presented with the help of tabular presentations. Although this critical survey cannot be collectively exhaustive, still this survey comprises the most significant works in the concerned paradigm by examining more than 300 research publications.

In this work, a new fractionally delayed biorthogonal wavelet is designed for recognition of voltage sag causes. This work addresses the classification of voltage sag causes into three categories, i.e., fault events (namely line-to-ground fault, line-to-line fault, double-line-to-ground fault and symmetrical fault), induction motor starting and transformer energization. Fractionally delayed biorthogonal Coiflet wavelet of order 2 (named as Coiflet fraclet) is designed using Lagrange interpolation and employed for the multiresolution analysis of voltage sag signals. To achieve higher classification accuracy, event information is derived from both time-frequency domain and frequency domain. Frequency information is obtained by applying multiple signal classification (MUSIC) algorithm on voltage sag signals. Both the time-frequency domain and frequency domain features are concatenated to implement the feature level fusion for strengthening the feature set. The feature set obtained after feature level fusion is used for training the recursive reduced kernel based extreme learning machine which addresses the problem of highly complex structure due to huge dataset of large dimensions. The classifier has shown robustness to the presence of noise also and classified the voltage sag causes with high accuracy.

This paper introduces a technique for designing an optimized matched wavelet by fractionally delaying the coefficients of the matched filter (FDMW). The method is an extension to Matched Wavelets Design process where the filter coefficients are statistically matched to the non-stationary discrete data. Filter coefficients obtained for such unit delay matched filters (UDW) have been fractionally delayed. Fractionally delayed filter coefficients are used to design Matched Wavelet followed by computation of the energy compaction ratio. The results show a significant improvement in the compression using FDMW over UDW. Designed FDM wavelets have lower energy compaction ratio and higher compression rate than given Unit Delay Wavelets.