Picked from the software, topic, and level of “Class Wavelet Power Quality Disturbance Classification #3: Discrete Wavelet Transform: Multi-Level Signal Decomposition”.
At most one email a week. Unfollow anytime with one click.
Transcript
The full narration of this video. The current sentence is highlighted while the video plays; click any sentence to jump there.
Show transcript
Introduction
Here is the final result, the transient signal is broken into six frequency bands with wave-flat. The fast oscillation appears clearly in the detail coefficients of level one exactly at the time of occurrence, something that Fourier transform cannot do. Let's start. Hello, welcome to episode three of the wave-flat series on power quality disturbance classification. Fourier transform tells us which frequencies are present, but not when they occur. Wave-flat breaks the signal into frequency bands while still preserving time information, making it suitable for short-duration disturbances like transients and sec. Today we are using the disc-rata wave-flat transform with pi-wave-flats.
Step 1: db4 decomposition at five levels
Create a file named dekomposisi.py The function wafedage breaks the signal with wave-flat using db4 up to 5 levels. Each level splits the signal into two, low-frequency approximation and high-frequency detail, then the approximation is split again. The result is six coefficients, A5 and D5 to D1. With a sampling frequency of 3.2 kHz, D1 contains 800 to 1600 Hz, D2 400 to 800, and so on until A5 below 50 Hz. We print the length and energy of each coefficient.
Step 2: Energy of each frequency band
Run. The coefficient length becomes shorter at deeper levels, 323 for D1 down to 26 for A5, because each level resamples half of it. On the sec signal, Almost all the energy is in A5 because its fundamental frequency is 50 Hz, and D1 is almost empty. In the transient signal, the energy in D1 increases more than 20 times, because the transient oscillation is in the high-frequency band.
Step 3: Sag & transients in wavelet coefficients
Now the graph. In sec, changes in amplitude are visible in A5 and D5, and the start and end edges of sec appear as small spikes in the details. In the transient, the coefficients of D1 and D2 remain constant over time except exactly when the oscillation occurs. This is the power of Wavelet, disturbance is localized in time and frequency, and this pattern will be our feature.
Done!
Done. We have understood the Wavelet Transform strata, 5-level decomposition with a decimation factor of 4, the frequency band of each coefficient, and observing localized disturbances in the detail coefficients. In the next episode, we will extract energy and statistical features from the Wavelet coefficients. Currently working on a thesis, dissertation, or research in the field of power quality and signal processing? Cerdas Inovasi Digital helps implement machine learning from generating disturbance signals according to IEEE-1159, Wavelet transformation, Feature extraction, up to classification with SVM, Random Forest, and XGBoost, as well as opening Python programming courses for electrical engineering. Want to master the entire series, from synthetic signals up to AIS testing? All 6 episodes, complete with step-by-step PDF guides and source code, are available for premium tutorial subscribers. Subscribe starting at Rp69.000 per month, save 28% with the 12-month package, or pay once Rp999.000 for lifetime access. Want just one class? Buy single classes with lifetime access. Subscribe now at the premium page cerdasinovasidigital.com Contact our WhatsApp, or visit cerdasinovasidigital.com Thank you for watching, see you in the next episode.
Ask the class material
Type a question or keyword; we find the part of this episode that explains it and jump the video there.
My notes
Notes are saved with the current video time and are only visible to you.
Questions & answers
Stuck on a step? Ask here; the instructor and other learners can answer.
No questions yet.
Your instructor
Cerdas Inovasi Digital Team
Software developers, BIM & 3D engineers, and AI practitioners
52 video classes1,276 episodes
We build software, BIM models, games and AI automations for clients every day, and turn that hands-on work into step-by-step classes: every step is shown on screen and comes with a PDF guide you can keep.
