English
Tiếng Việt
(en)A method, apparatus and computer readable medium for estimating a frequency of an electrical signal are disclosed. The method comprises taking a number of samples of an electrical signal throughout a main sampling window; dividing the number of samples taken throughout the main sampling window into a plurality of groups of consecutive samples, each group of consecutive samples forming a subwindow, wherein each subwindow includes a minimum number of the number of samples required to provide a required level of accuracy for estimating the fundamental frequency of the electrical signal; estimating, for each subwindow, the fundamental frequency of the electrical signal based on the group of consecutive samples of the respective subwindow; and determining if the fundamental frequency is stable throughout the main window by comparing the estimated fundamental frequency of each subwindow.
1.ApplicationNumber: US-201314653811-A
1.PublishNumber: US-2015355249-A1
2.Date Publish: 20151210
3.Inventor: CHANEDEAU GWLADYS FABIENNE ALICE
GAUBERT JEAN-PAUL
TISSIER JEAN-FRANCOIS
4.Inventor Harmonized: CHANEDEAU GWLADYS FABIENNE ALICE(FR)
GAUBERT JEAN-PAUL(FR)
TISSIER JEAN-FRANCOIS(FR)
5.Country: US
6.Claims:
(en)A method, apparatus and computer readable medium for estimating a frequency of an electrical signal are disclosed. The method comprises taking a number of samples of an electrical signal throughout a main sampling window; dividing the number of samples taken throughout the main sampling window into a plurality of groups of consecutive samples, each group of consecutive samples forming a subwindow, wherein each subwindow includes a minimum number of the number of samples required to provide a required level of accuracy for estimating the fundamental frequency of the electrical signal; estimating, for each subwindow, the fundamental frequency of the electrical signal based on the group of consecutive samples of the respective subwindow; and determining if the fundamental frequency is stable throughout the main window by comparing the estimated fundamental frequency of each subwindow.
7.Description:
(en)FIELD OF INVENTION
This disclosure relates to the field of signal analysis. More specifically, but not exclusively, the disclosure relates to a method and corresponding apparatus for analysing an electrical signal in order to determine a fundamental frequency of the signal along with associated harmonics.
BACKGROUND TO THE INVENTION
Stability of electricity grids is extremely important because so much of everyday life relies upon a stable and constant supply of electricity. If electricity supplied by an electricity grid is unstable then there is a high risk of a blackout occurring, which in some circumstances can result in large areas of a country, or even continent, losing power for relatively long periods of time. While blackouts are a general annoyance to everyone in a blackout zone, they also have an economic effect on businesses in the blackout zone and beyond.
In order to help maintain a stable electricity grid it is necessary to precisely monitor the electricity being supplied by the electricity grid in order to detect unusual fluctuations in the electricity supply, particularly in the frequency of the supply.
In recent years, electricity grids have started to become far more complex, particularly as more small scale renewable energy power sources are being integrated into electricity grids. Fortunately, new ranges of smart meters may help with such problems because the smart meters themselves are able to monitor certain characteristics of an electricity grid and feed such information back to a central server. Consequently, the central server is provided with more information from more sources, which makes the electricity grid as a whole better placed to keep a track of fluctuations in the electricity supply across the grid.
The resources available on smart meters for performing electricity signal analysis are limited and therefore many algorithms used by HV or MV substations to perform such frequency analysis are not suitable for use in a smart meter.
Simple methods of fundamental frequency estimation using zero crossings are well-known. However, such methods only obtain frequency information and do not obtain valuable information such as signal amplitude, and harmonic content of the electrical signal. Obtaining frequency, amplitude and harmonic information is required for obtaining a detailed understanding of the current state of the grid, throughout the grid. However, obtaining such detailed information to sufficient accuracy levels using the limited resources provided by smart meters is very difficult.
SUMMARY OF INVENTION
Embodiments of the present invention attempt to mitigate at least some of the above-mentioned problems.
In accordance with an aspect of the invention there is provided a method for estimating a frequency of an electrical signal, the method comprising taking a number of samples of an electrical signal throughout a main sampling window. The method also comprises dividing the number of samples taken throughout the main sampling window into a plurality of groups of consecutive samples, each group of consecutive samples forming a subwindow, wherein each subwindow includes a minimum number of the number of samples required to provide a required level of accuracy for estimating the fundamental frequency of the electrical signal. In addition, the method comprises estimating, for each subwindow, the fundamental frequency of the electrical signal based on the group of consecutive samples of the respective subwindow. The method also comprises determining if the fundamental frequency is stable throughout the main window by comparing the estimated fundamental frequency of each subwindow.
The method may further comprise determining a size of a main sampling window in terms of a number of samples prior to taking the number of samples.
In addition, the method may further comprise estimating the fundamental frequency over the main window if the fundamental frequency is determined to be stable throughout the main window.
The method may also further comprise performing harmonic analysis on the number of samples if the fundamental frequency is determined to be stable.
Furthermore, the method may further comprise determining a total harmonic distortion associated with the number of samples and re-sampling the electrical signal if the total harmonic distortion exceeds a threshold.
The fundamental frequency may be estimated using a maximum frequency point method. The fundamental frequency may be estimated using the maximum frequency point method in accordance with the following equation:
Max[ A (1)]=Max√{square root over ( a 1 2 ( F 1 )+ b 1 2 ( F 1 ))}{square root over ( a 1 2 ( F 1 )+ b 1 2 ( F 1 ))}
Wherein Max[A(1)] is the maximum magnitude value of fundamental frequency, F1 is the estimated fundamental frequency, and a1 and b1 are the fundamental discrete Fourier coefficients defined by:
a
1
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
cos
(
2
π
F
1
T
e
)
,
and
b
1
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
sin
(
2
π
F
1
T
e
)
Wherein T e is the sampling period, N is the total number of samples, N T is the total number of periods, i is the reference of a sample, and S(i) is a sample of the signal.
The fundamental frequency may be estimated over the main window by using the maximum frequency point method with a starting frequency equal to the average frequency over all the sub-windows.
Each subwindow may comprise approximately 700 samples. The main window may comprise at least 1200 samples. The main window preferably comprises approximately 5000 samples.
According to another aspect of the invention an apparatus is provided that is arranged to implement the method as described herein. The apparatus may be an electricity meter. 4. The electricity meter may further comprise a sensor arranged to take the samples, a memory arranged to store the one or more samples and store a computer program for implementing the method, and a processor arranged to perform the method in accordance with the stored samples and the computer program.
According to yet another aspect of the invention a computer readable medium is provided comprising computer readable code operable, in use, to instruct a computer to perform the method as described herein.
Embodiments of the invention provide a methodology for analysis of electrical signals including an accurate determination of the fundamental frequency of signals over successive sliding windows thereafter authorizing a reliable and significant harmonic analysis over all of the successive sliding windows.
Furthermore, embodiments of the invention provide an algorithmic principle utilizing Maximum Frequency Point (MFP) techniques as disclosed in PCT/FR2011/051575 to certify the stability of the fundamental frequency over a signal sampling window.
In certain embodiments of the invention algorithmic principles are used to decompose global voltage and current waveforms into their harmonic components at electricity meters.
Embodiments of the invention perform harmonic decomposition of electrical waveforms (voltage and current) based on the accurate knowledge of a fundamental frequency of the electrical waveform.
An algorithm is provided by embodiments of the invention that allows for control of the fundamental frequency during a studied window to provide stability both in terms of magnitude and phase. The extraction of the characteristics of the signal, such as magnitude and phase, for each harmonic component is possible with high accuracy. Furthermore, the stability of the magnitude of each harmonic component can also be verified by the algorithm provided in accordance with embodiments of the invention.
Embodiments of the invention provide an algorithm that can be integrated inside firmware of electricity meters. Such algorithms can also be utilized in residential electricity meters having low processing capabilities. The algorithm may be able to be integrated into existing electricity meters without an increase in hardware cost.
In embodiments of the invention, full harmonic analysis on every metrological channel is achievable.
Embodiments of the invention provide a system in which data size can be reduced to a minimum and can be sent to a data centre.
The algorithm according to embodiments of the invention permits an accurate calculation of the fundamental component (frequency, magnitude and phase) and harmonic range (magnitude and phase) throughout the studied window.
Embodiments of the invention allow for a detailed characterization of an electrical signal during a studied window. Some of the information that can be obtained may include one or more of:
1) The fundamental frequency; 2) Stability of the fundamental frequency over a main sampling window; 3) Fundamental rms magnitude and true rms value; 4) Fundamental magnitude state (i.e. stable, not stable); 5) THD values; 6) Selective harmonic analysis (including magnitude and phase); 7) Harmonic components state (i.e. stable, not stable) 8) Synthesis and transmission of pertinent power quality data.
BRIEF DESCRIPTION OF THE DRAWINGS
Exemplary embodiments of the invention shall now be described with reference to the drawings in which:
FIG. 1 illustrates a process for estimating characteristics of an electrical signal, including a fundamental frequency and harmonic components;
FIG. 2 shows a variation of fundamental frequency over a total duration of a main window comprising 60000 samples at a sampling rate of 256 μs; and
FIG. 3 shows a variation of the estimated fundamental frequency over subwindows of 10 periods of the fundamental frequency corresponding to 780 samples, wherein 76 subwindows are analyzed over a main window of 15,36s.
Throughout the description and the drawings, like reference numerals refer to like parts.
Specific Description
A method, arranged for implantation on a smart meter, for estimating frequency characteristics of an electrical signal, including a fundamental frequency and harmonic components, shall now be described with reference to FIG. 1 .
At step S 1 a number of samples, N, of the electrical characteristics of the signal are taken throughout a main sampling window. The samples may be voltage or current measurements of the electrical signal. Consequently, a suitable sensor within the smart meter is utilized to take the samples. The size of the main window, in terms of number of samples, is selected to balance the advantage of increased accuracy if the window is larger and more samples are therefore taken, with respect to the disadvantage of the additional processing and time required if more samples are taken. The size of the main window is determined by the smart meter; however, it will be appreciated that the size of the main window may be pre-programmed into the smart meter or updated on the smart meter remotely.
Once all of the samples have been taken, the sampled data throughout the main window is divided into a number of successive and adjacent subwindows at step S 2 . For each subwindow, the fundamental frequency of the signal is estimated using a Maximum Frequency Point (MFP) method as disclosed in PCT/FR2011/051575 and as shown by step S 3 and discussed in detail below.
The principle of the MFP method is the calculation of the fundamental discrete Fourier coefficients a 1 and b 1 of the sampled signal, as shown by Equations 1 and 2 respectively:
a
1
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
cos
(
2
π
F
1
T
e
)
Equation
1
b
1
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
sin
(
2
π
F
1
T
e
)
Equation
2
Wherein T e is the sampling period, N is the total number of samples, N T is the total number of periods, i is the reference of a sample, S(i) is a sample of the signal, F 1 is the estimated fundamental frequency.
The coefficients a 1 and b 1 are therefore calculated for each subwindow forming part of the main window. Then from the determined coefficients a 1 and b 1 of each subwindow the magnitude of the fundamental frequency component for each subwindow can be calculated in accordance with Equation 3:
Max[ A (1)]=Max√{square root over ( a 1 2 ( F 1 )+ b 1 2 ( F 1 ))}{square root over ( a 1 2 ( F 1 )+ b 1 2 ( F 1 ))} Equation 3
Wherein Max[A(1)] is the maximum magnitude value of fundamental frequency, a1 and b1 are the fundamental discrete Fourier coefficients and F1 is the estimated fundamental frequency.
The MFP method enables for estimation of the fundamental frequency at 50 Hz or 60 Hz with an accuracy of thousandths of hertz for a window of 5000 samples at a sampling period of 256 μs. The accuracy of the fundamental frequency estimation depends on the total number of samples associated with the sampling frequency. Hence, over a window of 700 samples an accuracy of thousandths of hertz can be obtained. For example, when F1=50 Hz and Te=256 μs, a sub-window size of 700 samples is provided, which corresponds to 10 periods of the main frequency. Consequently, a minimum window size of approximately 700 samples equivalent to 10 periods is required in order to provide the required accuracy.
Once all of the samples are taken, it is then determined whether or not the frequency is stable across all of the main window at step S 4 . If the fundamental frequency estimation is constant over all of the successive subwindows then it is determined that the fundamental frequency is stable throughout the main window with the accuracy of thousandths of hertz. More specifically, the smart meter determines that the frequency is stable if there is not a variation in the frequency of greater than +/−1 mHz over all of the subwindows. If the frequency variation is greater than +/−1 mHz then it is determined that the fundamental frequency is not stable, the accuracy of the fundamental frequency over the all studied window is not valid, and the analysis is stopped. Consequently, if the frequency is not stable then steps S 1 to S 3 are repeated again. This process is repeated until it is determined that the fundamental frequency is stable. FIG. 2 shows frequency instability over a period of samples during a trapezoidal frequency profile, while FIG. 3 shows frequency instability over intervals of 700 samples corresponding to 10 periods of the fundamental frequency.
If the frequency is stable then the global frequency, i.e. the frequency across the whole of the main window is determined at step S 5 . The fundamental frequency across the whole of the main window is determined by applying the MFP algorithm on all of the samples of the main window, with a starting frequency used for the process equal to the averaging of each subwindow frequency estimation.
Once the fundamental frequency F 1 has been proved to be stable over the whole of the studied main window, a Total Harmonic Distortion (THD) of the signal throughout the main sampling window is determined at step S 6 .
There are two standard definitions for calculating the THD of a periodical waveform, as discussed below.
The ratio between the total harmonic content of the signal during the sampling window and the RMS value of the component at the fundamental frequency as determined during steps S 1 to S 5 can be calculated in accordance with Equation 4:
T
H
D
(
1
)
=
Σ
n
=
2
∞
X
rms
n
2
X
rms
1
*
100
(
%
)
Equation
4
Wherein X rms 1 is the fundamental component of the signal, and X rms n is the harmonic component of the signal at rank n, its frequency is n×F1.
Furthermore, the ratio between the harmonic content during the window and the true RMS value of the fundamental frequency of the signal as determined during steps S 1 to S 5 can be calculated in accordance with Equation 5:
T
H
D
(
2
)
=
Σ
n
=
2
∞
X
rms
n
2
Σ
n
=
1
∞
X
rms
n
2
Equation
5
Equations 4 and 5 are mathematically linked by the relation shown by Equation 6:
T
H
D
(
1
)
=
T
H
D
(
2
)
1
-
(
T
H
D
(
2
)
2
)
Equation
6
The calculation of one THD (THD (1) or THD (2) ) over a signal window is then sufficient to know the two values of THD using equation 6.
If the THD value as defined by either Equation 4 or Equation 5 exceeds a maximum limit value linked to the THD definition, the harmonic analysis at step S 7 is performed. The THD limit is set in accordance with the IEC 61000 standard, IEEE 519 (Recommended Practices and Requirements for Harmonic Control in Electric Power systems in the USA). If the THD value as defined by either Equation 4 or Equation 5 exceeds the maximum limit value linked to the THD definition a new sample window capture process can be performed by returning to step S 1 .
When the THD value exceeds the maximum limit value linked to the THD definition; magnitude and phase of each harmonic range can be calculated at step S 7 . This is calculated using the Fourier coefficients, a h and b h , that have already been determined for each sample. Consequently, the discrete Fourier coefficients a h and b h at harmonic range h are calculated by Equations 7 and 8:
a
h
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
cos
(
2
π
h
F
1
T
e
)
Equation
7
Wherein T e is the sampling period, N is the total number of samples, N T is the total number of periods, i is the reference of a sample, S(i) is a sample of the signal, h is harmonic range, and F 1 is the estimated fundamental frequency.
b
h
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
sin
(
2
π
h
F
1
T
e
)
Equation
8
Wherein T e is the sampling period, N is the total number of samples, N T is the total number of periods, i is the reference of a sample, S(i) is a sample of the signal, h is harmonic range, and F 1 is the estimated fundamental frequency.
From the Fourier coefficients, the amplitude of the harmonic components are calculated using Equation 9:
A h = √{square root over ( a h 2 +b h 2 )} Equation 9
Wherein A h is the amplitude of the harmonic, a h and b h are discrete Fourier coefficient.
Furthermore, the phase of the harmonic components is calculated using Equation 10:
φ
h
=
tan
-
1
(
a
h
b
h
)
Equation
10
Wherein φ h is the phase of the harmonic components a h and b h are discrete Fourier coefficient.
The harmonic analysis therefore comprises carrying out the same Fourier analysis used to determine and analyse the fundamental frequency. The harmonic analysis is carried out on all harmonic frequencies that are integer multiples of the fundamental frequency.
At this point, the fundamental frequency along with the phase and amplitude components of the harmonic components have been determined accurately and reliably.
In practice, the above-described method can be implemented on any suitable device on an electricity grid for determining frequency characteristics of an electricity signal on the grid. However, the above-described method is arranged so that it is suitable for being implemented on a smart meter or other device having limited processing capabilities. The smart meters or other devices on the grid system that obtain this information are arranged to transfer the information to a centralized data centre. All such smart meters, or other devices on the network, are arranged to perform their measurements and calculations in accordance with a reference time of the main study window which is t 0 . Hence, acquisition by a number of smart meters is synchronized in time for all the metrological channels such as voltage and current of the meter. In consequence, all of the components such as the fundamental frequency and harmonic phases are synchronised to the same time reference, t 0 , within a single meter and across the network.
When a centralized data centre collects information of harmonic content from a number of electricity meters spread across a grid, the data centre is then able to perform harmonic pollution flow analysis. Hence, the collected information can be used by the data centre to monitor the grid in detail. Furthermore, the data centre can then use this information to control the grid by controlling smart meters or even MV/HV substations.
When implemented by a processor on a smart-meter or as a computer program on any type of computer, a computer would be provided having a memory to store the computer program, and a processor to implement the computer program. The processor would then perform the algorithmic process. The computer program may include computer code arranged to instruct a computer to perform the functions of one or more of the various methods described above. The computer program and/or the code for performing such methods may be provided to an apparatus, such as a computer, on a computer readable medium. The computer readable medium could be, for example, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, or a propagation medium for data transmission, for example for downloading the code over the Internet. Non-limiting examples of a physical computer readable medium include semiconductor or solid state memory, magnetic tape, a removable computer diskette, a random access memory (RAM), a read-only memory (ROM), a rigid magnetic disc, and an optical disk, such as a CD-ROM, CD-R/W or DVD.
1.PublishNumber: US-2015355249-A1
2.Date Publish: 20151210
3.Inventor: CHANEDEAU GWLADYS FABIENNE ALICE
GAUBERT JEAN-PAUL
TISSIER JEAN-FRANCOIS
4.Inventor Harmonized: CHANEDEAU GWLADYS FABIENNE ALICE(FR)
GAUBERT JEAN-PAUL(FR)
TISSIER JEAN-FRANCOIS(FR)
5.Country: US
6.Claims:
(en)A method, apparatus and computer readable medium for estimating a frequency of an electrical signal are disclosed. The method comprises taking a number of samples of an electrical signal throughout a main sampling window; dividing the number of samples taken throughout the main sampling window into a plurality of groups of consecutive samples, each group of consecutive samples forming a subwindow, wherein each subwindow includes a minimum number of the number of samples required to provide a required level of accuracy for estimating the fundamental frequency of the electrical signal; estimating, for each subwindow, the fundamental frequency of the electrical signal based on the group of consecutive samples of the respective subwindow; and determining if the fundamental frequency is stable throughout the main window by comparing the estimated fundamental frequency of each subwindow.
7.Description:
(en)FIELD OF INVENTION
This disclosure relates to the field of signal analysis. More specifically, but not exclusively, the disclosure relates to a method and corresponding apparatus for analysing an electrical signal in order to determine a fundamental frequency of the signal along with associated harmonics.
BACKGROUND TO THE INVENTION
Stability of electricity grids is extremely important because so much of everyday life relies upon a stable and constant supply of electricity. If electricity supplied by an electricity grid is unstable then there is a high risk of a blackout occurring, which in some circumstances can result in large areas of a country, or even continent, losing power for relatively long periods of time. While blackouts are a general annoyance to everyone in a blackout zone, they also have an economic effect on businesses in the blackout zone and beyond.
In order to help maintain a stable electricity grid it is necessary to precisely monitor the electricity being supplied by the electricity grid in order to detect unusual fluctuations in the electricity supply, particularly in the frequency of the supply.
In recent years, electricity grids have started to become far more complex, particularly as more small scale renewable energy power sources are being integrated into electricity grids. Fortunately, new ranges of smart meters may help with such problems because the smart meters themselves are able to monitor certain characteristics of an electricity grid and feed such information back to a central server. Consequently, the central server is provided with more information from more sources, which makes the electricity grid as a whole better placed to keep a track of fluctuations in the electricity supply across the grid.
The resources available on smart meters for performing electricity signal analysis are limited and therefore many algorithms used by HV or MV substations to perform such frequency analysis are not suitable for use in a smart meter.
Simple methods of fundamental frequency estimation using zero crossings are well-known. However, such methods only obtain frequency information and do not obtain valuable information such as signal amplitude, and harmonic content of the electrical signal. Obtaining frequency, amplitude and harmonic information is required for obtaining a detailed understanding of the current state of the grid, throughout the grid. However, obtaining such detailed information to sufficient accuracy levels using the limited resources provided by smart meters is very difficult.
SUMMARY OF INVENTION
Embodiments of the present invention attempt to mitigate at least some of the above-mentioned problems.
In accordance with an aspect of the invention there is provided a method for estimating a frequency of an electrical signal, the method comprising taking a number of samples of an electrical signal throughout a main sampling window. The method also comprises dividing the number of samples taken throughout the main sampling window into a plurality of groups of consecutive samples, each group of consecutive samples forming a subwindow, wherein each subwindow includes a minimum number of the number of samples required to provide a required level of accuracy for estimating the fundamental frequency of the electrical signal. In addition, the method comprises estimating, for each subwindow, the fundamental frequency of the electrical signal based on the group of consecutive samples of the respective subwindow. The method also comprises determining if the fundamental frequency is stable throughout the main window by comparing the estimated fundamental frequency of each subwindow.
The method may further comprise determining a size of a main sampling window in terms of a number of samples prior to taking the number of samples.
In addition, the method may further comprise estimating the fundamental frequency over the main window if the fundamental frequency is determined to be stable throughout the main window.
The method may also further comprise performing harmonic analysis on the number of samples if the fundamental frequency is determined to be stable.
Furthermore, the method may further comprise determining a total harmonic distortion associated with the number of samples and re-sampling the electrical signal if the total harmonic distortion exceeds a threshold.
The fundamental frequency may be estimated using a maximum frequency point method. The fundamental frequency may be estimated using the maximum frequency point method in accordance with the following equation:
Max[ A (1)]=Max√{square root over ( a 1 2 ( F 1 )+ b 1 2 ( F 1 ))}{square root over ( a 1 2 ( F 1 )+ b 1 2 ( F 1 ))}
Wherein Max[A(1)] is the maximum magnitude value of fundamental frequency, F1 is the estimated fundamental frequency, and a1 and b1 are the fundamental discrete Fourier coefficients defined by:
a
1
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
cos
(
2
π
F
1
T
e
)
,
and
b
1
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
sin
(
2
π
F
1
T
e
)
Wherein T e is the sampling period, N is the total number of samples, N T is the total number of periods, i is the reference of a sample, and S(i) is a sample of the signal.
The fundamental frequency may be estimated over the main window by using the maximum frequency point method with a starting frequency equal to the average frequency over all the sub-windows.
Each subwindow may comprise approximately 700 samples. The main window may comprise at least 1200 samples. The main window preferably comprises approximately 5000 samples.
According to another aspect of the invention an apparatus is provided that is arranged to implement the method as described herein. The apparatus may be an electricity meter. 4. The electricity meter may further comprise a sensor arranged to take the samples, a memory arranged to store the one or more samples and store a computer program for implementing the method, and a processor arranged to perform the method in accordance with the stored samples and the computer program.
According to yet another aspect of the invention a computer readable medium is provided comprising computer readable code operable, in use, to instruct a computer to perform the method as described herein.
Embodiments of the invention provide a methodology for analysis of electrical signals including an accurate determination of the fundamental frequency of signals over successive sliding windows thereafter authorizing a reliable and significant harmonic analysis over all of the successive sliding windows.
Furthermore, embodiments of the invention provide an algorithmic principle utilizing Maximum Frequency Point (MFP) techniques as disclosed in PCT/FR2011/051575 to certify the stability of the fundamental frequency over a signal sampling window.
In certain embodiments of the invention algorithmic principles are used to decompose global voltage and current waveforms into their harmonic components at electricity meters.
Embodiments of the invention perform harmonic decomposition of electrical waveforms (voltage and current) based on the accurate knowledge of a fundamental frequency of the electrical waveform.
An algorithm is provided by embodiments of the invention that allows for control of the fundamental frequency during a studied window to provide stability both in terms of magnitude and phase. The extraction of the characteristics of the signal, such as magnitude and phase, for each harmonic component is possible with high accuracy. Furthermore, the stability of the magnitude of each harmonic component can also be verified by the algorithm provided in accordance with embodiments of the invention.
Embodiments of the invention provide an algorithm that can be integrated inside firmware of electricity meters. Such algorithms can also be utilized in residential electricity meters having low processing capabilities. The algorithm may be able to be integrated into existing electricity meters without an increase in hardware cost.
In embodiments of the invention, full harmonic analysis on every metrological channel is achievable.
Embodiments of the invention provide a system in which data size can be reduced to a minimum and can be sent to a data centre.
The algorithm according to embodiments of the invention permits an accurate calculation of the fundamental component (frequency, magnitude and phase) and harmonic range (magnitude and phase) throughout the studied window.
Embodiments of the invention allow for a detailed characterization of an electrical signal during a studied window. Some of the information that can be obtained may include one or more of:
1) The fundamental frequency; 2) Stability of the fundamental frequency over a main sampling window; 3) Fundamental rms magnitude and true rms value; 4) Fundamental magnitude state (i.e. stable, not stable); 5) THD values; 6) Selective harmonic analysis (including magnitude and phase); 7) Harmonic components state (i.e. stable, not stable) 8) Synthesis and transmission of pertinent power quality data.
BRIEF DESCRIPTION OF THE DRAWINGS
Exemplary embodiments of the invention shall now be described with reference to the drawings in which:
FIG. 1 illustrates a process for estimating characteristics of an electrical signal, including a fundamental frequency and harmonic components;
FIG. 2 shows a variation of fundamental frequency over a total duration of a main window comprising 60000 samples at a sampling rate of 256 μs; and
FIG. 3 shows a variation of the estimated fundamental frequency over subwindows of 10 periods of the fundamental frequency corresponding to 780 samples, wherein 76 subwindows are analyzed over a main window of 15,36s.
Throughout the description and the drawings, like reference numerals refer to like parts.
Specific Description
A method, arranged for implantation on a smart meter, for estimating frequency characteristics of an electrical signal, including a fundamental frequency and harmonic components, shall now be described with reference to FIG. 1 .
At step S 1 a number of samples, N, of the electrical characteristics of the signal are taken throughout a main sampling window. The samples may be voltage or current measurements of the electrical signal. Consequently, a suitable sensor within the smart meter is utilized to take the samples. The size of the main window, in terms of number of samples, is selected to balance the advantage of increased accuracy if the window is larger and more samples are therefore taken, with respect to the disadvantage of the additional processing and time required if more samples are taken. The size of the main window is determined by the smart meter; however, it will be appreciated that the size of the main window may be pre-programmed into the smart meter or updated on the smart meter remotely.
Once all of the samples have been taken, the sampled data throughout the main window is divided into a number of successive and adjacent subwindows at step S 2 . For each subwindow, the fundamental frequency of the signal is estimated using a Maximum Frequency Point (MFP) method as disclosed in PCT/FR2011/051575 and as shown by step S 3 and discussed in detail below.
The principle of the MFP method is the calculation of the fundamental discrete Fourier coefficients a 1 and b 1 of the sampled signal, as shown by Equations 1 and 2 respectively:
a
1
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
cos
(
2
π
F
1
T
e
)
Equation
1
b
1
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
sin
(
2
π
F
1
T
e
)
Equation
2
Wherein T e is the sampling period, N is the total number of samples, N T is the total number of periods, i is the reference of a sample, S(i) is a sample of the signal, F 1 is the estimated fundamental frequency.
The coefficients a 1 and b 1 are therefore calculated for each subwindow forming part of the main window. Then from the determined coefficients a 1 and b 1 of each subwindow the magnitude of the fundamental frequency component for each subwindow can be calculated in accordance with Equation 3:
Max[ A (1)]=Max√{square root over ( a 1 2 ( F 1 )+ b 1 2 ( F 1 ))}{square root over ( a 1 2 ( F 1 )+ b 1 2 ( F 1 ))} Equation 3
Wherein Max[A(1)] is the maximum magnitude value of fundamental frequency, a1 and b1 are the fundamental discrete Fourier coefficients and F1 is the estimated fundamental frequency.
The MFP method enables for estimation of the fundamental frequency at 50 Hz or 60 Hz with an accuracy of thousandths of hertz for a window of 5000 samples at a sampling period of 256 μs. The accuracy of the fundamental frequency estimation depends on the total number of samples associated with the sampling frequency. Hence, over a window of 700 samples an accuracy of thousandths of hertz can be obtained. For example, when F1=50 Hz and Te=256 μs, a sub-window size of 700 samples is provided, which corresponds to 10 periods of the main frequency. Consequently, a minimum window size of approximately 700 samples equivalent to 10 periods is required in order to provide the required accuracy.
Once all of the samples are taken, it is then determined whether or not the frequency is stable across all of the main window at step S 4 . If the fundamental frequency estimation is constant over all of the successive subwindows then it is determined that the fundamental frequency is stable throughout the main window with the accuracy of thousandths of hertz. More specifically, the smart meter determines that the frequency is stable if there is not a variation in the frequency of greater than +/−1 mHz over all of the subwindows. If the frequency variation is greater than +/−1 mHz then it is determined that the fundamental frequency is not stable, the accuracy of the fundamental frequency over the all studied window is not valid, and the analysis is stopped. Consequently, if the frequency is not stable then steps S 1 to S 3 are repeated again. This process is repeated until it is determined that the fundamental frequency is stable. FIG. 2 shows frequency instability over a period of samples during a trapezoidal frequency profile, while FIG. 3 shows frequency instability over intervals of 700 samples corresponding to 10 periods of the fundamental frequency.
If the frequency is stable then the global frequency, i.e. the frequency across the whole of the main window is determined at step S 5 . The fundamental frequency across the whole of the main window is determined by applying the MFP algorithm on all of the samples of the main window, with a starting frequency used for the process equal to the averaging of each subwindow frequency estimation.
Once the fundamental frequency F 1 has been proved to be stable over the whole of the studied main window, a Total Harmonic Distortion (THD) of the signal throughout the main sampling window is determined at step S 6 .
There are two standard definitions for calculating the THD of a periodical waveform, as discussed below.
The ratio between the total harmonic content of the signal during the sampling window and the RMS value of the component at the fundamental frequency as determined during steps S 1 to S 5 can be calculated in accordance with Equation 4:
T
H
D
(
1
)
=
Σ
n
=
2
∞
X
rms
n
2
X
rms
1
*
100
(
%
)
Equation
4
Wherein X rms 1 is the fundamental component of the signal, and X rms n is the harmonic component of the signal at rank n, its frequency is n×F1.
Furthermore, the ratio between the harmonic content during the window and the true RMS value of the fundamental frequency of the signal as determined during steps S 1 to S 5 can be calculated in accordance with Equation 5:
T
H
D
(
2
)
=
Σ
n
=
2
∞
X
rms
n
2
Σ
n
=
1
∞
X
rms
n
2
Equation
5
Equations 4 and 5 are mathematically linked by the relation shown by Equation 6:
T
H
D
(
1
)
=
T
H
D
(
2
)
1
-
(
T
H
D
(
2
)
2
)
Equation
6
The calculation of one THD (THD (1) or THD (2) ) over a signal window is then sufficient to know the two values of THD using equation 6.
If the THD value as defined by either Equation 4 or Equation 5 exceeds a maximum limit value linked to the THD definition, the harmonic analysis at step S 7 is performed. The THD limit is set in accordance with the IEC 61000 standard, IEEE 519 (Recommended Practices and Requirements for Harmonic Control in Electric Power systems in the USA). If the THD value as defined by either Equation 4 or Equation 5 exceeds the maximum limit value linked to the THD definition a new sample window capture process can be performed by returning to step S 1 .
When the THD value exceeds the maximum limit value linked to the THD definition; magnitude and phase of each harmonic range can be calculated at step S 7 . This is calculated using the Fourier coefficients, a h and b h , that have already been determined for each sample. Consequently, the discrete Fourier coefficients a h and b h at harmonic range h are calculated by Equations 7 and 8:
a
h
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
cos
(
2
π
h
F
1
T
e
)
Equation
7
Wherein T e is the sampling period, N is the total number of samples, N T is the total number of periods, i is the reference of a sample, S(i) is a sample of the signal, h is harmonic range, and F 1 is the estimated fundamental frequency.
b
h
=
1
N
T
∑
i
=
0
N
-
1
2
T
e
T
1
S
(
)
sin
(
2
π
h
F
1
T
e
)
Equation
8
Wherein T e is the sampling period, N is the total number of samples, N T is the total number of periods, i is the reference of a sample, S(i) is a sample of the signal, h is harmonic range, and F 1 is the estimated fundamental frequency.
From the Fourier coefficients, the amplitude of the harmonic components are calculated using Equation 9:
A h = √{square root over ( a h 2 +b h 2 )} Equation 9
Wherein A h is the amplitude of the harmonic, a h and b h are discrete Fourier coefficient.
Furthermore, the phase of the harmonic components is calculated using Equation 10:
φ
h
=
tan
-
1
(
a
h
b
h
)
Equation
10
Wherein φ h is the phase of the harmonic components a h and b h are discrete Fourier coefficient.
The harmonic analysis therefore comprises carrying out the same Fourier analysis used to determine and analyse the fundamental frequency. The harmonic analysis is carried out on all harmonic frequencies that are integer multiples of the fundamental frequency.
At this point, the fundamental frequency along with the phase and amplitude components of the harmonic components have been determined accurately and reliably.
In practice, the above-described method can be implemented on any suitable device on an electricity grid for determining frequency characteristics of an electricity signal on the grid. However, the above-described method is arranged so that it is suitable for being implemented on a smart meter or other device having limited processing capabilities. The smart meters or other devices on the grid system that obtain this information are arranged to transfer the information to a centralized data centre. All such smart meters, or other devices on the network, are arranged to perform their measurements and calculations in accordance with a reference time of the main study window which is t 0 . Hence, acquisition by a number of smart meters is synchronized in time for all the metrological channels such as voltage and current of the meter. In consequence, all of the components such as the fundamental frequency and harmonic phases are synchronised to the same time reference, t 0 , within a single meter and across the network.
When a centralized data centre collects information of harmonic content from a number of electricity meters spread across a grid, the data centre is then able to perform harmonic pollution flow analysis. Hence, the collected information can be used by the data centre to monitor the grid in detail. Furthermore, the data centre can then use this information to control the grid by controlling smart meters or even MV/HV substations.
When implemented by a processor on a smart-meter or as a computer program on any type of computer, a computer would be provided having a memory to store the computer program, and a processor to implement the computer program. The processor would then perform the algorithmic process. The computer program may include computer code arranged to instruct a computer to perform the functions of one or more of the various methods described above. The computer program and/or the code for performing such methods may be provided to an apparatus, such as a computer, on a computer readable medium. The computer readable medium could be, for example, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, or a propagation medium for data transmission, for example for downloading the code over the Internet. Non-limiting examples of a physical computer readable medium include semiconductor or solid state memory, magnetic tape, a removable computer diskette, a random access memory (RAM), a read-only memory (ROM), a rigid magnetic disc, and an optical disk, such as a CD-ROM, CD-R/W or DVD.
You are contracting for Fundamental frequency stability and harmonic analysis
Expert Fundamental frequency stability and harmonic analysis
You are commenting for Fundamental frequency stability and harmonic analysis