I was going through the ways in which it has been solved in real life and on the work that has been done upon Power Line Fault, being an electrical engineer we always deals with harmonics and since transmission lines over long distances add noisy harmonics to the signals due to the small but significant capacitance and inductance in the transmission lines.
\n\nThus there are noise always in the signal, but the fault also creates transients just as the transients we observe while we switch on our water heater and our lighting bulbs or tubelights in our house flickkers for a moment, this happens because the transients get generated and signals with different frequencies and phase add up to make a wave which has low frequency and thus our eye is able to spot changes.
\n\nSo going by the intuition we have noise transients frequencies and fault transients frequency, band pass type filter should be made to weed out noices to make a process signal for further analysis for fault transients.
\n\nI came accross the following research paper, they are working on the same problem and they have used wavelet transform to extract features and built a SVM model to classify. Although the research paper is published in 2011, but its a good read to give domain knowledge and basic direction
\n\n", "messages": [ { "id": "443664", "postDate": "12/22/2018 04:29:00", "content": "I was going through the ways in which it has been solved in real life and on the work that has been done upon Power Line Fault, being an electrical engineer we always deals with harmonics and since transmission lines over long distances add noisy harmonics to the signals due to the small but significant capacitance and inductance in the transmission lines.
\n\nThus there are noise always in the signal, but the fault also creates transients just as the transients we observe while we switch on our water heater and our lighting bulbs or tubelights in our house flickkers for a moment, this happens because the transients get generated and signals with different frequencies and phase add up to make a wave which has low frequency and thus our eye is able to spot changes.
\n\nSo going by the intuition we have noise transients frequencies and fault transients frequency, band pass type filter should be made to weed out noices to make a process signal for further analysis for fault transients.
\n\nI came accross the following research paper, they are working on the same problem and they have used wavelet transform to extract features and built a SVM model to classify. Although the research paper is published in 2011, but its a good read to give domain knowledge and basic direction
\n\n", "rawMarkdown": "I was going through the ways in which it has been solved in real life and on the work that has been done upon Power Line Fault, being an electrical engineer we always deals with harmonics and since transmission lines over long distances add noisy harmonics to the signals due to the small but significant capacitance and inductance in the transmission lines. \n\nThus there are noise always in the signal, but the fault also creates transients just as the transients we observe while we switch on our water heater and our lighting bulbs or tubelights in our house flickkers for a moment, this happens because the transients get generated and signals with different frequencies and phase add up to make a wave which has low frequency and thus our eye is able to spot changes.\n\nSo going by the intuition we have noise transients frequencies and fault transients frequency, band pass type filter should be made to weed out noices to make a process signal for further analysis for fault transients.\n\nI came accross the following research paper, they are working on the same problem and they have used wavelet transform to extract features and built a SVM model to classify. Although the research paper is published in 2011, but its a good read to give domain knowledge and basic direction\n\nhttps://www.researchgate.net/publication/251998942_Transmission_line_fault_detection_and_classification", "votes": null }, { "id": "444818", "postDate": "12/24/2018 22:35:09", "content": "Hi Nukkda, thanks for your sharing. I am wondering what's the sample dimension of those CWT data?
", "rawMarkdown": "Hi Nukkda, thanks for your sharing. I am wondering what's the sample dimension of those CWT data?", "votes": null }, { "id": "447553", "postDate": "12/30/2018 04:00:13", "content": "Section 2.1 denoising and signal preprocessing of the following paper also discusses the use of DWT on signals: http://dspace.vsb.cz/bitstream/handle/10084/133114/VAN431_FEI_P1807_1801V001_2018.pdf
", "rawMarkdown": "Section **2.1 denoising and signal preprocessing** of the following paper also discusses the use of DWT on signals: http://dspace.vsb.cz/bitstream/handle/10084/133114/VAN431_FEI_P1807_1801V001_2018.pdf", "votes": null }, { "id": "447554", "postDate": "12/30/2018 04:03:10", "content": "I found this short video on wavelets useful https://www.youtube.com/watch?v=QX1-xGVFqmw
", "rawMarkdown": "I found this short video on wavelets useful https://www.youtube.com/watch?v=QX1-xGVFqmw", "votes": null }, { "id": "448121", "postDate": "12/31/2018 09:29:28", "content": "The problem is that at what frequencies these transients are. And note that the host said their oscilloscopes are cheap and only sample at 40 MHz. He said PD activities may not be captured in the typical way.
", "rawMarkdown": "The problem is that at what frequencies these transients are. And note that the host said their oscilloscopes are cheap and only sample at 40 MHz. He said PD activities may not be captured in the typical way.", "votes": null } ], "comments": [ { "id": 444818, "author_name": "zuoxiaozhen", "author_url": "", "post_date": "12/24/2018 22:35:09", "content": "Hi Nukkda, thanks for your sharing. I am wondering what's the sample dimension of those CWT data?
", "votes": null, "replies": [] }, { "id": 447553, "author_name": "jackvial", "author_url": "", "post_date": "12/30/2018 04:00:13", "content": "Section 2.1 denoising and signal preprocessing of the following paper also discusses the use of DWT on signals: http://dspace.vsb.cz/bitstream/handle/10084/133114/VAN431_FEI_P1807_1801V001_2018.pdf
", "votes": null, "replies": [] }, { "id": 447554, "author_name": "jackvial", "author_url": "", "post_date": "12/30/2018 04:03:10", "content": "I found this short video on wavelets useful https://www.youtube.com/watch?v=QX1-xGVFqmw
", "votes": null, "replies": [] }, { "id": 448121, "author_name": "zoujie", "author_url": "", "post_date": "12/31/2018 09:29:28", "content": "The problem is that at what frequencies these transients are. And note that the host said their oscilloscopes are cheap and only sample at 40 MHz. He said PD activities may not be captured in the typical way.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "443664": "I was going through the ways in which it has been solved in real life and on the work that has been done upon Power Line Fault, being an electrical engineer we always deals with harmonics and since transmission lines over long distances add noisy harmonics to the signals due to the small but significant capacitance and inductance in the transmission lines. \n\nThus there are noise always in the signal, but the fault also creates transients just as the transients we observe while we switch on our water heater and our lighting bulbs or tubelights in our house flickkers for a moment, this happens because the transients get generated and signals with different frequencies and phase add up to make a wave which has low frequency and thus our eye is able to spot changes.\n\nSo going by the intuition we have noise transients frequencies and fault transients frequency, band pass type filter should be made to weed out noices to make a process signal for further analysis for fault transients.\n\nI came accross the following research paper, they are working on the same problem and they have used wavelet transform to extract features and built a SVM model to classify. Although the research paper is published in 2011, but its a good read to give domain knowledge and basic direction\n\nhttps://www.researchgate.net/publication/251998942_Transmission_line_fault_detection_and_classification", "444818": "Hi Nukkda, thanks for your sharing. I am wondering what's the sample dimension of those CWT data?", "447553": "Section **2.1 denoising and signal preprocessing** of the following paper also discusses the use of DWT on signals: http://dspace.vsb.cz/bitstream/handle/10084/133114/VAN431_FEI_P1807_1801V001_2018.pdf", "447554": "I found this short video on wavelets useful https://www.youtube.com/watch?v=QX1-xGVFqmw", "448121": "The problem is that at what frequencies these transients are. And note that the host said their oscilloscopes are cheap and only sample at 40 MHz. He said PD activities may not be captured in the typical way." }, "source": "meta" }