This paper presents a new convolutional neural network (CNN) topology using wavelet kernels to detect and discriminate single or multiple Partial Discharge (PD) locations in high voltage power apparatus with increased accuracy. The method is tested on an electrical equipment model with acoustic PD sensors. A cubical tank has been emulated in the laboratory representing the equipment under test and partial discharge sources have been placed at different positions along with the required data acquisition hardware. The present scheme eliminates the requirement of separate algorithms for feature extraction and classification of the acquired PD signals. Wavelet kernels of the CNN play a crucial role in feature learning, and the proposed CNN architecture as a whole, can classify the features in a supervised manner. The performance of the proposed scheme is compared with other existing methods using the same dataset. It is found that an overall accuracy of 97.64% is achieved by the proposed method, outperforming other existing methods by a significant margin of at least 5% in terms of accuracy. The developed module is a generic one and can be adapted for different high voltage electrical apparatus with similar topological structures; hence it can be used in various ways in power industry.
The development of the electric power grid has boosted the use of metal oxide surge arresters in electrical systems. These devices serve as protective measures and aid in insulation coordination. Monitoring the condition of surge arresters is crucial to ensure the reliability of the power system. Among various techniques, conditional monitoring based on leakage current signals, particularly resistive leakage current analysis is popular and effective. In this work, the resistive leakage current has been separated from the measured total leakage current signal and several features have been extracted to train the machine learning model for suitable identification of the contamination level of the surge arresters.
Electric spring is a new concept in arena of conventional AC Power system. In comparison with conventional STATCOM, FACTS equipment, Electric Spring has an edge over both active and reactive power controllability. Introduction of solid state devices in power system provokes harmonic disturbance and affects equipment performance. Hence it is also very crucial to ensure to control harmonics presence in power system. This paper delve in delve into a new mechanism of ES for faster operation and better THD.
1 June 2025 in International Journal of Applied Power Engineering (IJAPE)
The present work represents a method for identification of the vulnerable nodes in smart grid as well as assessment of the performance of voltage stability indicator technique with the help of weighted least square scheme. in today’s smart grid system, false data injection (FDI) is the major issue to supply uninterruptedly at demand side in advanced metering infrastructure (AMI). The recent blackouts are the consequence of non-identifying FDI as research on FDI is not considered under power system analysis. In our research, vulnerable nodes of a power system network have been identified and a state estimation method was used to eliminate superfluous data for those identified nodes. Voltage stability indicator (VSI) based state estimation have been used successfully to make the smart grid system error free as possible. VSI method has been used first to find the vulnerable nodes of the grid after that the efficient state estimation method i.e. optimal weighted least square (optimal WLS) have been employed to get refined result. Results show that VSI based technique in concurrence with optimal WLS has potential to eliminate undesirable data with sensible level of precision.
1 January 2026 in IEEE Transactions on Dielectrics and Electrical Insulation