Assessment of Errors of Fundamental Frequency Estimation Methods in the Presence of Voltage Fluctuations and Distortions
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Electrical Engineering. The day in electrical engineering: signals, systems and the hardware that carries them.
Rosa: We're looking at the paper: "Assessment of Errors of Fundamental Frequency Estimation Methods in the Presence of Voltage Fluctuations and Distortions".
Dev: Here's the gist of the paper. The paper presents numerical simulation studies to assess errors in various fundamental frequency estimation methods when analyzing voltage signals with fluctuations and distortions, which is crucial for power quality diagnostics The dispersion of fundamental frequency estimation errors shows that none of the methods allow for error values for the analyzed signal to be lower than those required by the relevant standard The best results are obtained for the IEC method and the modified ESPRIT method Pruning redundant reasoning tokens The research was conducted in the Matlab programming environment where a test signal was numerically generated to recreate typical disturbances in a low-voltage power network and for which the fundamental frequency at specific times is known This test signal can be described by the relationship utest(t) =
1 + kAMumod(t): ·X H h-1 Uamph cos 2πhf0t + φh + h∆f02π umod(t)max Z umod(t)dt + noise The signal is a model of simultaneously occurring signal fluctuations and distortions for which it is possible to unambiguously determine the fundamental frequency The harmonic values for the studies were assumed to be those for a “clipped cosine” signal
9: for various clipped level mc (mc = 0.01, 0.8, 1) The assumed modulating signal has step changes because in practice load variability often occurs abruptly due to the sudden disconnection/connection of loads How the training signal is built The test signal utest(t) is defined by equation (7) as umod(t) = sign(sin (2πfmt)) where fm is the modulating frequency The base values of the fundamental frequency f0 (carrier frequency) were assumed to be 47, 49.98, 50, 50.02, and 52 Hz respectively The frequency deviations ∆f0 were assumed to be 0.1, 1, and 10 Hz respectively The study analyzes signals without noise and with noise at SNR levels of 10, 0, and-10 dB respectively Where the gains come from The analysis shows that increasing the modulating frequency fm and deviation ∆f0 causes an increase in the fundamental frequency estimation errors for all methods considered Different levels of noise and distortion have little influence on the estimation results of individual methods The modified ESPRIT method provided lowest error values Both the IEC method and the modified ESPRIT method provide the best results and are characterize by fundamental frequency estimation errors reduced to several percent The research presented in this paper indicates the need to develop a new effective method for fundamental frequency estimation and to verify the measuring and recording equipment using the test signal specified in the paper It is important to carry out metrological evaluation of existing measuring and recording devices by using the test signal indicated in the paper which recreates simultaneous fluctuations and distortions The paper concludes that none of the methods allows for error values for the analyzed signal to be lower than those required by the relevant standard It is worth noting that a significant increase in estimation errors occurs with increasing frequency changes of the fundamental frequency of the measured signal and the level of variability of this frequency The paper also suggests it is necessary to looking for new methods of estimating the fundamental frequency which will guarantee obtaining results characterized by an error smaller than that defined in the relevant standard Pruning redundant reasoning tokens The IEC 61000-4-30 standard specifies that the fundamental frequency should be determined according to the following relationship fe0 = NT0 / ∆tNT0 The standard specifies that to minimize the impact of multiple zero crossings, harmonics and interharmonics should be suppressed before each estimation For the purposes of the study, a 100th-order band-pass filter with finite impulse response was adopted Pruning redundant reasoning tokens The autocorrelation method determines the fundamental frequency as the inverse of the fundamental period estimated by the method The fundamental period using this method is determined as the average value of the difference in the local maxima of the autocorrelation function in the time domain taking into account information about the sampling frequency For a more stable operation of the method when determining local maxima it is assumed that the distance between local maxima cannot be less than approximately 13.3 ms Pruning redundant reasoning tokens The Hilbert method determines the fundamental frequency for a given time interval as the average value of the instantaneous frequency finst(t) in the analyzed window The instantaneous frequency finst(t) is determined according to the relationship finst(t) = 1 / 2π dφ(t)/dt where φ(t) is the instantaneous phase determined as the argument of the analytical signal of the measured signal Pruning redundant reasoning tokens The ESPRIT algorithm models the sampled waveform as a superposition of complex exponentials given a sequence of sampled data x(n) of length L Defining hk = Ake jψk can be rewritten in matrix form as xb = VΦnH + r The damping factors and frequencies are estimated from the real and imaginary parts of the natural logarithm of eigenvalues λbi respectively Once the frequencies and damping factors are known the amplitudes and initial phases can be estimated using a least-squares approach The traditional sliding-window ESPRIT method provides accurate spectral estimates but suffers from high computational complexity Pruning redundant reasoning tokens The dispersion of fundamental frequency estimation errors δf0 in the form of boxplot grouped by the modulating frequency fm for the individual methods considered is shown in Fig.
Rosa: 3 Analyzing the obtained results it can be seen that for all the methods considered increasing the modulating frequency fm and deviation ∆f0 causes an increase in the fundamental frequency estimation errors Pruning redundant reasoning tokens The dispersion of fundamental frequency estimation errors δf0 in the form of boxplot grouped by the clipped level mc for the individual methods considered is shown in Fig.
Dev: 6 The modified ESPRIT method provided lowest error values However the IEC method has lower computational complexity The best and most stable results were obtained for the IEC method although at certain modulating frequencies The modified ESPRIT method provided lowest error values Pruning redundant reasoning tokens The dispersion of fundamental frequency estimation errors δf0 in the form of boxplot grouped by the SNR level for the individual methods considered is shown in Fig.
Title and authors: Rosa: So we're looking at this paper called "Assessment of Errors of Fundamental Frequency Estimation Methods in the Presence of Voltage Fluctuations and Distortions." It’s by Antonio Bracale, Pasquale De Falco, Piotr Kuwałek, and Grzegorz Wiczyński. This paper tackles how accurately we can find the fundamental frequency when the voltage signal is actually messy with fluctuations and distortions.
Dev: Yeah, so it’s about making sure we get this parameter—the fundamental frequency—right in modern power grids because getting it wrong can cause real diagnostic problems. It uses numerical simulations to test different ways of finding that frequency under realistic, nonstationary conditions.
Taro: And the big takeaway seems to be that none of the methods they tested can guarantee an error lower than what relevant standards require for this type of signal <ref:2601.16606#pg2>. So they’re showing us where we fall short when things get complicated.
Rosa: Exactly, because they used a test signal specifically designed to mimic the kinds of simultaneous voltage fluctuations and distortions you see in real low-voltage power networks <ref:2601.16606#pg3>. They built this signal to be really tough on any estimation technique.
Dev: That’s what they did, generating a signal where the fundamental frequency is known at specific times, which lets them test everything systematically. They used various harmonic levels too, like a clipped cosine with levels of zero point zero one and zero point eight or one <ref:2601.16606#pg3>.
Rosa: And they varied things like how fast the signal is fluctuating, calling that modulating frequency fm, and how much that frequency drifts from the expected value, which they call the deviation delta f naught <ref:2601.16606#pg3>.
Taro: So increasing both of those factors—the modulation speed and the frequency drift—actually caused these fundamental frequency estimation errors to increase for every single method they looked at <ref:2601.16606#pg3>.
Dev: It’s a bit worrying then, because it suggests that the more dynamic and distorted the signal gets, the harder it is for our current tools to keep up with the required accuracy. But they did find some winners.
Rosa: They did, and they pointed to two specific approaches as being better: the standard IEC method and a modified version of ESPRIT <ref:2601.16606#pg3>. The modified ESPRIT method gave them the lowest error values overall in those simulations.
Dev: The paper notes that even with these best results, both the IEC method and the modified ESPRIT method keep their fundamental frequency estimation errors reduced to just a few percent <ref:2601.16606#pg2>. That's pretty tight for real-world use.
Title and authors: Taro: So for someone who only cares about what this means practically, it tells us that we need to seriously consider switching from the standard IEC approach if we’re dealing with these kinds of fluctuating signals.
Rosa: Right, and they also looked at how noise levels affected things—they tested signal-to-noise ratios at ten decibels, zero decibels, and even negative ten decibels <ref:2601.16606#pg3>.
Dev: And interestingly, the paper found that the level of noise or distortion didn't have much influence on the estimation results for any single method; it just made things a little noisier overall <ref:2601.16606#pg3>.
Taro: That’s interesting because it means we can focus our improvement efforts more on the signal itself, like its frequency changes, instead of getting bogged down in cleaning up the noise first.
Rosa: Precisely, and they highlighted that the modified ESPRIT method was particularly good at handling those distortions with low errors <ref:2601.16606#pg3>. This suggests that complex estimation algorithms are necessary for this kind of power quality analysis.
Dev: They also mentioned how the IEC sixty-one thousand-four-thirty standard works, which says to minimize issues from multiple zero crossings and harmonics before estimating the fundamental frequency <ref:2601.16606#pg2>. They used a hundredth order bandpass filter with finite impulse response for that in their study <ref:2601.16606#pg3>.
Taro: It sounds like they are really pushing us to think about pre-processing the signal before we even try to estimate the frequency, which is something I hadn't thought about much in a real-time system context.
Rosa: And when you look at the methods they tested—like autocorrelation, Hilbert method, and ESPRIT—the ESPRIT algorithm is known for being computationally complex because it has to solve an eigenvalue problem for every sliding window <ref:2601.16606#pg3>.
Dev: That complexity is a huge issue for real-time loops where you need fast responses and low latency; if the loop rate isn't high enough, that traditional ESPRIT method just won't work in practice <ref:2601.16606#pg3>.
Taro: So they proposed a modified scheme in another study to cut down on that computational cost while still keeping the accuracy up for things like harmonic distortion <ref:2601.16606#pg3>. That’s the kind of engineering trade-off we have to navigate every day.
Rosa: The paper suggests that the need is more urgent than ever to develop a new, more effective method for fundamental frequency estimation, one that can guarantee an error smaller than what those standards demand <ref:2601.16606#pg3>.
Title and authors: Dev: Because right now, based on their tests with this specific test signal, none of the existing methods allow you to get the required error values down to the necessary level <ref:2601.16606#pg2>. So we’re looking at a gap in our current diagnostic toolkit.
Taro: It sounds like they are laying out a clear path for future research, moving beyond just testing existing tools and actually building something new that can handle the real mess of power grids <ref:2601.16606#pg3>.
Rosa: And they also emphasize that we need to carry out metrological evaluations of our existing measuring and recording equipment using this specific test signal <ref:2601.16606#pg3>. It’s a direct call for us to test our hardware under the exact conditions they simulated.
Dev: So, to wrap up the main findings of "Assessment of Errors of Fundamental Frequency Estimation Methods in the Presence of Voltage Fluctuations and Distortions," we see that increasing signal variability makes estimation errors grow across all techniques <ref:2601.16606#pg3>.
Taro: And while they found better performance with the modified ESPRIT method, we also see that the IEC method has lower computational complexity, which is a trade-off we have to consider <ref:2601.16606#pg3>.
Rosa: Ultimately, this paper suggests that significant increases in estimation errors occur when you change the frequency of the signal or increase its variability <ref:2601.16606#pg3>. We need new methods to get those error values down to the standard level they set <ref:2601.16606#pg3>.
Dev: So, for our listeners, this means that any system relying on fundamental frequency estimation needs to be re-evaluated when it encounters dynamic power quality issues <ref:2601.16606#pg2>. It’s not just about a clean signal anymore.
Taro: I think the biggest implication is that we need to look for fundamentally new ways of estimating this frequency, not just tweaking the parameters of these existing algorithms <ref:2601.16606#pg3>.
Rosa: Exactly, and they really urge us to use this test signal for our own equipment evaluations before we deploy anything in a real power grid setting <ref:2601.16606#pg3>. We’re talking about proactive testing based on these kinds of realistic disturbances.
Dev: So, to wrap up the work on "Assessment of Errors of Fundamental Frequency Estimation Methods in the Presence of Voltage Fluctuations and Distortions," the conclusion is that we need new methods if we want to meet those strict standards <ref:2601.16606#pg2>.
Taro: I just think the focus needs to shift toward building methods that inherently handle these simultaneous fluctuations and distortions gracefully <ref:2601.16606#pg3>.
Rosa: That’s the direction they are pointing us in, moving toward methods that are robust against real power grid conditions <ref:2601.16606#pg3>. We’ll be looking for those new ideas next time.
The paper's summary: Rosa: So, to wrap up what we just heard about these methods failing under stress, this paper is looking at how well different ways of finding a fundamental frequency work when the voltage signal is actually fluctuating and distorted <ref:2601.16606#pg3>.
Dev: Yeah, it’s essentially a numerical test to see if any standard way of estimating that frequency can keep up with real power grid problems, like when you have those sudden load changes and noise messing things up.
Rosa: The main finding they report is pretty blunt: none of the methods they tested can produce an error value for the analyzed signal that is lower than what the relevant industry standards require <ref:2601.16606#pg2>.
Dev: That’s a tough message for diagnostic engineers because it suggests that if you're relying on one of these standard approaches, you can't guarantee you'll meet the required accuracy under these kinds of messy conditions.
Rosa: They found that things get worse as the signal gets more dynamic—the modulation frequency increases and the deviation of the fundamental frequency drifts further away from its expected value <ref:2601.16606#pg3>.
Dev: And they did show some winners, though. The modified ESPRIT method came out on top in terms of giving the lowest error values for that test signal <ref:2601.16606#pg3>.
Rosa: But even with those best results, both the standard IEC method and that modified ESPRIT method are still keeping their fundamental frequency estimation errors down to just a few percent <ref:2601.16606#pg2>. That's what you’re getting when you push the system to its limits.
Dev: It’s interesting because they also looked at how noise levels affect this, and they found that the level of noise or distortion didn't really change how much error each individual method made <ref:2601.16606#pg3>. That means the problem isn't just the background hiss, it's the signal itself being too volatile.
Rosa: Exactly, and this leads them to a pretty clear conclusion: they need to develop a completely new way of estimating that fundamental frequency if they want results that actually meet those strict standards <ref:2601.16606#pg3>.
Dev: So what this means for us on the ground is that we can’t just keep using the same tools blindly when you see these kinds of severe fluctuations in your measurements; you need a different approach entirely if you want to be metrologically sound.
Rosa: And they even push for everyone to use this specific test signal to do their own evaluations of their measuring and recording equipment, so we can really see how our hardware holds up when the world misbehaves <ref:2601.16606#pg3>.
Dev: It’s a call for proactive testing; not just checking if your device works in a quiet lab environment, but seeing how it handles these specific kinds of simultaneous fluctuations and distortions that happen out in the field.
Rosa: So, moving forward, the focus isn't just on tweaking the existing algorithms like ESPRIT or IEC; it’s about building something fundamentally new that is robust against those real power grid conditions <ref:2601.16606#pg3>.
The paper's improvements: Rosa: So, after seeing how much error these current methods make under dynamic conditions, the paper isn't just stopping there; they’re pointing toward a real need for better estimation techniques <ref:2601.16606#pg3>.
Dev: Yeah, they are basically saying that we can’t keep just tweaking the parameters of these existing algorithms like ESPRIT or IEC and expect them to magically hit those strict industry standards <ref:2601.16606#pg3>.
Rosa: The main suggestion is to develop a brand new method for fundamental frequency estimation that is inherently better at handling simultaneous fluctuations and distortions gracefully <ref:2601.16606#pg3>.
Dev: That's a big ask because developing something completely new takes way more time and effort than just optimizing the existing mathematical operators like lambda or sigma <ref:2601.16606#pg3>.
Rosa: And they really want us to start testing our own measuring and recording equipment with this specific test signal so we can see exactly how well our hardware performs when the world gets messy <ref:2601.16606#pg3>.
Dev: That's crucial for the safety case, because if our equipment doesn't handle these kinds of real-world disturbances well, then any system relying on its fundamental frequency will be unreliable <ref:2601.16606#pg2>.
Rosa: They are pushing us to move beyond just checking if a device works in a quiet lab; they want us to evaluate it under these kinds of realistic, nonstationary power grid conditions <ref:2601.16606#pg3>.
Dev: It means the future work is going to be about finding methods that are robust against those kinds of frequency changes and high variability that we saw cause the errors to spike in the first place <ref:2601.16606#pg3>.
Rosa: Right, so it’s not just a software fix; it’s a call for new research into estimation techniques that are designed from the ground up to survive real power quality issues <ref:2601.16606#pg3>.
Dev: And that new method needs to be fast enough for real-time loops, or else the whole point of having an estimator is lost when you’re dealing with high loop rates <ref:2601.16606#pg3>.
Conclusion: Rosa: So we’ve covered how different methods perform when the voltage signal gets really messy, that's what this paper "Assessment of Errors of Fundamental Frequency Estimation Methods in the Presence of Voltage Fluctuations and Distortions" is all about <ref:2601.16606#pg3>.
Dev: Yeah, they basically show us that none of the current methods can guarantee meeting the required accuracy for this kind of signal, no matter which one you pick <ref:2601.16606#pg2>.
Rosa: The big picture is that we really need to develop a completely new way to estimate the fundamental frequency if we want results that actually meet those strict standards <ref:2601.16606#pg3>.
Dev: And they are pushing us to use this test signal for our own equipment evaluations so we can see how our hardware holds up when the world gets misbehaved <ref:2601.16606#pg3>.
Taro: I just think the focus needs to shift toward building methods that are robust against those kinds of simultaneous fluctuations and distortions gracefully <ref:2601.16606#pg3>.
Rosa: Exactly, moving toward something that doesn't break when the environment gets volatile <ref:2601.16606#pg3>.
Dev: I agree, and that new method needs to be fast enough for real-time loops or the whole point of having an estimator is lost in a slow system <ref:2601.16606#pg3>.
Taro: And that new approach has to handle those frequency changes and high variability we saw cause the errors to spike in the first place <ref:2601.16606#pg3>.
Rosa: So, even with the best results they found for the modified ESPRIT method, it still shows a significant increase in estimation errors when things get dynamic <ref:2601.16606#pg3>.
Dev: It highlights that we can’t just rely on existing tools without a serious plan for how they will perform under real power quality issues <ref:2601.16606#pg2>.
Taro: The implications are that we need new research into estimation techniques that are designed from the ground up to survive these kinds of real power grid conditions <ref:2601.16606#pg3>.
Rosa: So, in summary, this paper is a strong call for developing better methods if we want to move past just tweaking existing algorithms <ref:2601.16606#pg3>.
Dev: We’ll keep an eye on those new ideas as they come out and see if they can actually handle the latency issues we deal with in real-time systems <ref:2601.16606#pg3>.
Taro: Definitely, because the reliability of our diagnostic tools depends on what happens when the system misbehaves, not just when it runs smoothly <ref:2601.16606#pg3>.
Antonio Bracale, Pasquale De Falco, Piotr Kuwałek, Grzegorz Wiczy łnski
University of Naples Parthenope · Institute of Electrical Engineering and Electronics, Pozna University of Technology
eess.SP
Submitted: 2026-01-23
Updated: 2026-11-07
Comments: 5 pages, 7 figures, submitted to IEEE conferences: 2026 22nd International Conference on Harmonics & Quality of Power (ICHQP)
Journal ref: A. Bracale, P. De Falco, P. Kuwałek and G. Wiczyński, Assessment of Errors of Fundamental Frequency Estimation Methods in the Presence of Voltage Fluctuations and Distortions, 2026 22nd ICHQP, Dresden, Germany, 2026
DOI: 10.1109/ICHQP67927.2026.11756967
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 72/100
The gist: The paper presents numerical simulation studies to assess errors in various fundamental frequency estimation methods when analyzing voltage signals with fluctuations and distortions…
Key concepts
- Test Signal utest(t)
- A numerically generated signal designed to mimic real power network disturbances, including fluctuations and distortions. It allows researchers to test estimation methods under realistic, non-ideal conditions where the true fundamental frequency is known.
- Modulating Frequency fm and Deviation Delta f0
- These parameters define the characteristics of the signal's fluctuations. Increasing these values causes a measurable increase in the errors produced by all tested frequency estimation methods, indicating that high variability makes accurate measurement difficult.
- Modified ESPRIT Method
- A specific algorithm used for frequency estimation that performed better than others in this study. It provided the lowest error values when analyzing signals affected by noise and distortion, suggesting it is a strong candidate for improved diagnostic tools.
Terminology
Summary
The paper presents numerical simulation studies to assess errors in various fundamental frequency estimation methods when analyzing voltage signals with fluctuations and distortions, which is crucial for power quality diagnostics The dispersion of fundamental frequency estimation errors shows that none of the methods allow for error values for the analyzed signal to be lower than those required by the relevant standard The best results are obtained for the IEC method and the modified ESPRIT method
Pruning redundant reasoning tokens
The research was conducted in the Matlab programming environment where a test signal was numerically generated to recreate typical disturbances in a low-voltage power network and for which the fundamental frequency at specific times is known This test signal can be described by the relationship utest(t) = [1 + kAMumod(t)]·X H h-1 Uamph cos 2πhf0t + φh + h∆f02π umod(t)max Z umod(t)dt + noise The signal is a model of simultaneously occurring signal fluctuations and distortions for which it is possible to unambiguously determine the fundamental frequency The harmonic values for the studies were assumed to be those for a “clipped cosine” signal [9] for various clipped level mc (mc = 0.01, 0.8, 1) The assumed modulating signal has step changes because in practice load variability often occurs abruptly due to the sudden disconnection/connection of loads
How the training signal is built
The test signal utest(t) is defined by equation (7) as umod(t) = sign(sin (2πfmt)) where fm is the modulating frequency The base values of the fundamental frequency f0 (carrier frequency) were assumed to be 47, 49.98, 50, 50.02, and 52 Hz respectively The frequency deviations ∆f0 were assumed to be 0. p. 2 02, and 52 Hz, respectively. The frequency deviations ∆f0 were assumed to be 0.Where the paper says it
Where the gains come from
The analysis shows that increasing the modulating frequency fm and deviation ∆f0 causes an increase in the fundamental frequency estimation errors for all methods considered Different levels of noise and distortion have little influence on the estimation results of individual methods The modified ESPRIT method provided lowest error values Both the IEC method and the modified ESPRIT method provide the best results and are characterize by fundamental frequency estimation errors reduced to several percent
Conclusion
The research presented in this paper indicates the need to develop a new effective method for fundamental frequency estimation and to verify the measuring and recording equipment using the test signal specified in the paper It is important to carry out metrological evaluation of existing measuring and recording devices by using the test signal indicated in the paper which recreates simultaneous fluctuations and distortions The paper concludes that none of the methods allows for error values for the analyzed signal to be lower than those required by the relevant standard It is worth noting that a significant increase in estimation errors occurs with increasing frequency changes of the fundamental frequency of the measured signal and the level of variability of this frequency The paper also suggests it is necessary to looking for new methods of estimating the fundamental frequency which will guarantee obtaining results characterized by an error smaller than that defined in the relevant standard
The IEC 61000-4-30 standard specifies that the fundamental frequency should be determined according to the following relationship fe0 = NT0 / ∆tNT0 The standard specifies that to minimize the impact of multiple zero crossings, harmonics and interharmonics should be suppressed before each estimation For the purposes of the study, a 100th-order band-pass filter with finite impulse response was adopted
The autocorrelation method determines the fundamental frequency as the inverse of the fundamental period estimated by the method The fundamental period using this method is determined as the average value of the difference in the local maxima of the autocorrelation function in the time domain taking into account information about the sampling frequency For a more stable operation of the method when determining local maxima it is assumed that the distance between local maxima cannot be less than approximately 13.3 ms
The Hilbert method determines the fundamental frequency for a given time interval as the average value of the instantaneous frequency finst(t) in the analyzed window The instantaneous frequency finst(t) is determined according to the relationship finst(t) = 1 / 2π dφ(t)/dt where φ(t) is the instantaneous phase determined as the argument of the analytical signal of the measured signal
The ESPRIT algorithm models the sampled waveform as a superposition of complex exponentials given a sequence of sampled data x(n) of length L Defining hk = Ake jψk can be rewritten in matrix form as xb = VΦnH + r The damping factors and frequencies are estimated from the real and imaginary parts of the natural logarithm of eigenvalues λbi respectively Once the frequencies and damping factors are known the amplitudes and initial phases can be estimated using a least-squares approach The traditional sliding-window ESPRIT method provides accurate spectral estimates but suffers from high computational complexity
The dispersion of fundamental frequency estimation errors δf0 in the form of boxplot grouped by the modulating frequency fm for the individual methods considered is shown in Fig. 3 Analyzing the obtained results it can be seen that for all the methods considered increasing the modulating frequency fm and deviation ∆f0 causes an increase in the fundamental frequency estimation errors
The dispersion of fundamental frequency estimation errors δf0 in the form of boxplot grouped by the clipped level mc for the individual methods considered is shown in Fig. 6 The modified ESPRIT method provided lowest error values However the IEC method has lower computational complexity The best and most stable results were obtained for the IEC method although at certain modulating frequencies The modified ESPRIT method provided lowest error values
The dispersion of fundamental frequency estimation errors δf0 in the form of boxplot grouped by the SNR level for the individual methods considered is shown in Fig.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Develop an AI system capable of assessing errors in fundamental frequency estimation methods by comparing their outputs against a known test signal that recreates
simultaneous fluctuations and distortions.
This allows for a direct evaluation of method accuracy under realistic, nonstationary power grid conditions. -
Create an AI system that can select the optimal fundamental frequency estimation method (IEC 61000-4-30 Method, Autocorrelation method, Hilbert method, Modified ESPRIT method) based on predefined performance metrics derived from simulations of varying noise levels (SNR) and distortion types (clipped level mc).
-
Implement an AI system that can predict the resulting fundamental frequency estimation error delta f0 given specific input signal characteristics such as the modulating frequency fm and the deviation of fundamental frequency Delta f0, allowing for proactive identification of problematic operational regimes.
-
Enhance power quality diagnostics by enabling the AI system to identify when any existing method fails to meet the required accuracy defined in standards like IEC 61000-4-30, as the paper concludes that
none of the methods allows for error values for the analyzed signal to be lower than those required by the relevant standard.
-
Improve fault detection capabilities by using an AI system trained on this test signal to specifically detect disturbances associated with high frequency changes and high variability, as
a significant increase in estimation errors occurs with increasing frequency changes of the fundamental frequency of the measured signal and the level of variability of this frequency.
Abstract
The fundamental frequency is one of the parameters that define power quality. Correctly determining this parameter under the conditions that prevail in modern power grids is crucial. Diagnostic purposes often require an efficient estimation of this parameter within short time windows. Therefore, this article presents the results of numerical simulation studies that allow the assessment of errors in various fundamental frequency estimation methods, including the standard IEC 61000-4-30 method, when the analyzed signal has a form similar to that found in modern power grids. For the purposes of this study, a test signal was adopted recreating the states of the power grid, including the simultaneous occurrence of voltage fluctuations and distortions. Conclusions are presented based on conducted research.
Related papers
- Open-H-RF: A Large-Scale Ultrasound Raw Channel Dataset Initiative
- Scalable Attention for 5G NR Channel Estimation
- Low-Altitude Wireless Networks: A Comprehensive Survey
- Sensing-Assisted OFDM with Index Modulation: A Waveform for Optical Wireless Integrated Communication and Localisation
- Iteration-Free Candidate Pruning and an Index-Domain VLSI Architecture for Sort-Free MIMO Box Decoding up to 1024-QAM
- A Foundation Model for Large-Scale Wireless Network Planning, Operation and Optimization