Detection of PQ Short Duration Variations using Stockwell Transform with LSTM

Ali, Mohamed Ali; Oda, Eyad S; Abdelsalam, Abdelazeem; Abdelaziz, Almoataz Y

doi:10.21608/sceee.2023.240048.1006

Detection of PQ Short Duration Variations using Stockwell Transform with LSTM

Document Type : Original Article

Authors

¹ Operation Engineer

² Electrical Engineering Department, Faculty of Engineering, Suez Canal University, Ismailia, Egypt

³ Ismailia

⁴ Electrical Power & Machines Dept., Faculty of Engineering, Ain Shams University, Cairo, Egypt

10.21608/sceee.2023.240048.1006

Abstract

Classification and detection of power quality disturbances (PQDs) are high priorities within the electrical power system. We are using feature extraction with artificial intelligence (AI) and deep learning to solve PQD problems using a two-step methodology: Feature extraction and classification steps, with the Feature extraction step utilizing Stockwell Transform and the classification step employing Long Short-Term Memory techniques. This work aims to use Stockwell Transform as a feature extraction using a Deep Learning (DL) approach known as LSTM for the classification and detection of PQ disturbance occurrences.
Signal characteristics are collected from the time-frequency analysis data based on Stockwell transform utilizing the Deep Learning technique in the long short-term memory (LSTM) network, which finds and classifies PQ disturbance events. By integrating the S-transform with the long short-term memory (LSTM) network, it is possible to achieve a high level of classification efficiency. Many PQ disturbances are treated with single and combination disruptions. The findings demonstrate that the proposed approach is precise and robust in detecting and identifying single and combination PQ disruptions. In comparison with many concise studies, the proposed strategy performs well.

Keywords