Apparatus and method for power detection and for generating power patterns
Patent Information
- Application Number
- AU2025257170
- Authority / Receiving Office
- AU · AU
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-04-18
- Filing Date
- 2025-04-17
- Publication Date
- 2026-10-29
Smart Images
Figure 00000000_0000_ABST
Abstract
Description
The invention relates to a device for determining power in electrical power systems as described in Patent Claim 1, as well as a method for doing so as described in Claim 16. Power measurement in multi-conductor circuits is defined in Germany by the DIN 40110-2 standard and worldwide by the IEE-1459 standard. The calculation methods specified in these standards form the basis for power calculations in modern measuring instruments. According to the standard, the specified operating range is ±1 Hz relative to a nominal frequency of 50 Hz and ±1.2 Hz relative to a nominal frequency of 60 Hz. Overtones and harmonics are taken into account. In the low-frequency range, they are only considered starting at 15 Hz and above. Nonlinear system behavior of the grid is thus not taken into account. Since an increasing number of decentralized inverter-based systems will be put into operation in the future—systems that influence grid behavior in the low-frequency range—the power calculation results will be distorted if the low-frequency range of the power-relevant variables, namely grid current and grid voltage, is not taken into account. The algorithms currently in use, based on Fourier analysis, produce errors for low-frequency modulated nominal frequencies that, depending on the technical implementation (sampling rate, window length), exceed permissible tolerance values by a wide margin. It should be noted that a Fast Fourier Transform (FFT) provides information in the frequency domain with a resolution that depends on the time window (measurement interval). No information about the signal itself can be reproduced. To circumvent this limitation, a “windowed” Fourier transform was introduced to generate a frequency-time spectrum. This variant of the FFT is also referred to as the SFT (Short-Time Fourier Transform). However, this short-time Fourier transform has the disadvantage that a fixed window width must be used. This characteristic has a significant impact on the frequency / time resolution and thus on the accuracy with which a signal can be identified. U.S. Patent No. 9,170,986 B2 discloses a power analysis method that determines information about harmonic components based on wavelets, in particular the reconstruction of waveforms that are analyzed and processed in other devices and applications. However, this method eliminates low-frequency components (sag / swell cycles), which, according to voltage quality standards (DIN-EN 50160), fall under the term “flicker.” The objective of the present invention is to provide a device and a method for determining power in power grids such that low-frequency signal components are also taken into account during power determination. In particular, the device is intended to enable the evaluation of transient nonlinear processes and to actively utilize these for the accurate and simultaneous, time-synchronized evaluation of power and energy signals. To solve this problem, the invention exhibits the features of Patent Claim 1. According to the invention, predetermined wavelets are used for fixed frequency intervals, with the frequency range extending from 0 Hz up to a maximum frequency. The frequency range to be analyzed is divided into frequency intervals in such a way that frequency intervals currently used in standards are respected, and the corresponding current and voltage values are applied to the wavelet function (wavelet pattern) in each case. The invention takes advantage of the fact that the length of the wavelets depends on the frequency interval under investigation or the frequency level. By dividing the current and voltage signals into predetermined frequency intervals, a frequency-based parallel analysis or calculation of power can be performed. In this way, highly accurate measurement of individual frequencies is ensured for all power components—in particular active and reactive power—as well as for the rated frequency. On the one hand, the invention enables precise power determination, whereby the exact amplitudes of the frequency components are determined in a first measurement path. On the other hand, an evaluation of grid disturbances can be performed based on power patterns, whereby the power components are determined in a second measurement path at a specific predetermined time point across a predetermined frequency range. The superposition of the transmitted power components forms power patterns that can characterize specific grid disturbances. This enables the provision of energy efficiency services that subsequently lead to the detection of a current disturbance in the grid. According to the invention, the first measurement path can be used for precise power measurement, and the second measurement path can be used for time-precise determination of power patterns. Alternatively, a combination of both measurement paths can be used to determine precise and simultaneously time-precise voltage and current values. According to a preferred embodiment of the invention, the temporal duration of the wavelet patterns varies across different predetermined frequency ranges. In a frequency range that encompasses relatively high frequencies—namely, for example, frequencies between 50 Hz and 4 kHz— and in which, for example, the current and voltage values are sampled at equidistant 25 Hz intervals, the wavelet patterns are shorter in duration than in another frequency range or frequency band, which lies between 1 Hz and 50 Hz, and in which, for example, the current and voltage values are sampled in equidistant 1 Hz frequency bands. Thus, depending on the selected frequency range or frequency band, there is a discontinuous change in the temporal length of the wavelet patterns. The temporal length of the wavelet patterns depends preferably on Heisenberg’s uncertainty principle, according to which a computation time is inversely proportional to the frequency. According to an alternative embodiment of the invention, the durations of the wavelet patterns may increase continuously and inversely proportional to the sampling frequency. However, the computational effort required for this is higher. According to a preferred embodiment of the invention, the wavelet pattern is configured as a Daubechies wavelet pattern, specifically as an orthogonal Daubechies wavelet pattern. In particular, if it is a Daubechies wavelet DB10 with 20 coefficients, the values of the scaling and wavelet functions are normalized to 1, so that they can be used directly as root mean square (RMS) values for determining current and voltage. This simplifies the determination of the effective power, since the frequency-selective current and voltage values can be multiplied together as rootmean-square (RMS) values and summed. This makes the determination of power or energy efficient; no additional conversion is required. Validation of the measured values is easily performed. In a further embodiment of the invention, the signals—namely, the grid current and the grid voltage—are sampled at equidistant frequency intervals. Advantageously, the amplitude and frequency component values are of equal magnitude. According to a further embodiment of the invention, a predetermined latency period— by which the wavelet values calculated in different frequency ranges (frequency intervals) are shifted—depends on a minimum frequency. Only after the latency period has elapsed can the signal be displayed and used for further processing. It is taken into account that wavelets have varying time and frequency resolutions. To represent or measure non-stationary signals, the observation period and frequency must be adapted to the signal being observed. In this regard, low-frequency signal components must be observed for a longer period, or a longer wavelet must be used. To determine the exact power, larger time intervals must therefore be used, typically in the range of minutes. This latency time specifies the maximum run-time correction required to synchronize the respective frequency-dependent or frequency-interval-dependent wavelet values. Using the frequency-based runtime correction module, a “reverse transformation” of the time point to the observation time point thus takes place—that is, the point from which one wishes to account for the exact shift of the wavelet signal relative to the useful signal (e.g., 50 Hz grid frequency). If, for example, the frequency interval is set to 1 Hz, the lowest frequency value can be set to 0.01 Hz, resulting in a latency of 100 seconds. After these 100 seconds have elapsed, the power values can be determined by summing the current and voltage signals present at the observation time within the respective frequency intervals (at the beginning of the 100 seconds) and applying the corresponding time shift to the power values at the observation time. In a further embodiment of the invention, a time-based runtime correction module is provided, which enables time-accurate measurement. This allows for the determination of patterns and time-accurate assignments of frequency components. For example, this allows for the detection of rapidly changing dynamics, such as aging processes, fault detection in machine parts, and changes in the operational behavior of systems and devices. The time-based runtime correction module ensures that all wavelet values are measured with precise timing. This enables the reconstruction of the original signal from the individual components. In a further embodiment of the invention, an AI module is provided in which fault pattern recognition algorithms are collected so that current faults can be detected by comparing them with the currently measured values. To solve this problem, the method exhibits the features of Patent Claim 16. Examples of the invention’s implementation are explained in more detail below with reference to the drawings. They show: Fig. 1 A block diagram of a measuring device according to the invention, Fig. 2 Processing of measured power and energy values, broken down by frequency components, Fig. 3 a graph showing a function of frequency or time intervals for different time windows based on the Heisenberg uncertainty principle, Fig. 4 a representation of wavelets with different resolutions, Fig. 5a a representation of two derived test signals, Fig. 5b an analysis of the first test signal using Fast Fourier Transform (FFIT) and wavelets (WT), Fig. 5c an analysis of the second test signal using Fast Fourier Transform (FFIT) and wavelets (WT), Fig. 6 a representation of the frequency ranges under investigation, Fig. 7a a representation of the determined frequency components of the wavelets for a specific disturbance in the power grid, Fig. 7b an enlarged view of a low-frequency component of the wavelet amplitudes shown in Figure 7a, Fig. 8 an example illustration of wavelet values at different frequencies (frequency components), Fig. 9 a representation of a Daubechies wavelet pattern (D20) and Fig. 10 a table of the wavelet coefficients of the Daubechies wavelet pattern shown in Figure 9 at an example sample frequency of fS = 7.680 Hz. A device according to the invention for determining power in a power grid 1 essentially consists of a current measuring device 2 for measuring a grid current at a measurement point in the power grid 1, a voltage measuring device 3 for measuring a grid voltage at the measurement point in the power grid 1, a sampling device 4 for sampling the measured grid current and grid voltage, and a processing device 5 for processing the sampled grid current and grid voltage values. Preferably, the device additionally comprises a communication / control device 6 that is connected via a communication interface 7 to remotely located communication participants. The communication interface 7 can be used, in particular, to transmit the determined power patterns that were recorded at a specific point in time. The determined power patterns can be used for control and protection applications—for example, via an algorithm—to, for instance, control a protective device or trigger protective measures. Preferably, the device includes a display 22 for displaying the determined power results. The processing unit 5 is divided into two sub-processing units 5a and 5b, wherein the respective sub-processing units 5a and 5b can be executed independently of one another. The first sub-processing unit 5a serves to determine an exact power and energy measurement. A second processing sub-unit 5b is used to rapidly detect changes and the corresponding timing of the individual frequency components for use, for example, in pattern recognition and error analysis (NILM—Non-Intrusive Load Monitoring) or predictive maintenance. The following figures illustrate the power and power components that can be determined using the device according to the invention. Apparent power S: S = ^P2 + Q2 Active power P: p2 (Pg + po + PM + Reactive power Q: Q2 (qg + Q0 + Qm 0 SM m=1 / + Do2 + Dm2 + DM02 00 0 + S (Dwmo2) + S (1^) + S (Dwm2) + U2 m=1 m=1 m=1 Total active power P in watts o PG: Active power component: fundamental frequency = 1 o Po: Active power component: distortion >1 o PM: Active power component: modulation 0< <1 o Pw: Active power component: interactions m e (1. to) 0 < v < 1 Total reactive power Q in var is composed of the components o Qg : Displacement reactive power: fundamental frequency v = 1 o Qo : Displacement reactive power: distortion v >1 o Qm : Displacement reactive power: modulation 0 < v < 1 o QwDisplacement reactive power: interactions m e (1 . ^)0 < y < 1 o D0: Deformation reactive power: distortion v >1 o Dm: Deformation reactive power: modulation 0 < v < 1 o Dwm: Deformation reactive power: modulation m e (1 . to)0 < y < 1 o Dmo: Interaction between Do and DM m e (1 . to) v >1 o D^ Interaction of DM and DWm m e (1. to) v > 1 o Day / j Interaction of Do and DWi l m e (1 . to) v > 1 ^: Asymmetric reactive power The underlined power components are not determined by conventional power measurement devices, but are unique to the measurement device of the invention. Figure 5a illustrates the derivation of test signals St1 and St2 from two different signals S1 and S2 of different frequencies. Figures 5b and 5c illustrate the analysis based on a Fast Fourier Transform (FFT) and using a wavelet transform, respectively. After the Fourier transform, frequency values of varying magnitudes are determined, whereas after the wavelet transform, identical wavelet values are determined in each case, regardless of the composition of the frequency components in the test signals St1 and St2, respectively. The signals S1 and S2 are harmonic oscillations, with the frequency of signal S1 being lower than that of signal S2. The first test signal St1 is obtained by superimposing or summing the first signal S1 and the second signal S2. The second test signal St2 is produced by the sequential repetition of the first signal S1 and the second signal S2. As can be seen in Figure 5a, the second test signal St2 consists exclusively of the first signal S1 with the lower frequency in a first time interval and exclusively of the signal S2 with the higher frequency in a second interval. If the Fast Fourier Transform (FFT) is applied to the first test signal St1, it follows from Figure 5b that the test signal St1 consists of two frequencies: a lower frequency f1 and a higher frequency f2. If the wavelet transform is applied to the first test signal St1, two values, W1 and W2, are obtained, where the first value, W1, corresponds to the lower frequency f1 and the second value, W2, corresponds to the higher frequency f2. These two constant values, W1 and W2, are generated over the entire time window. It should be noted that wavelets are wave patterns of a specific duration. Well-known examples include Morlet wavelets and Daubechies wavelets, which are suitable as pattern functions for power measurement. Figure 4, for example, shows a wavelet (wavelet function) WL. Within a time window of + / - 20 time units, it exhibits oscillations with an amplitude that rises in the first half and falls in the second half. With respect to time unit zero, the wavelet is symmetric; this means that the rising first half is symmetric to the falling second half of the wavelet and exhibits its maximum amplitude in magnitude at the center of the time window. When a Fast Fourier Transform (FFT) and a wavelet transform are applied to the second test signal St2, the result is shown in Figure 5c. Pursuant to the Fast Fourier Transform (FFT), a value corresponding to the lower frequency f1 is obtained in a first time window where the first signal S1, with lower frequency f1, is present, and a value corresponding to the higher frequency f2 is obtained in a second time interval where the signal S2, with higher frequency f2, is present. According to the wavelet transform, the value W1, which corresponds to the lower frequency f1, is obtained in the first time interval T1, and the value W2, which corresponds to the higher frequency f2, is obtained in the second time interval T2. In addition, the wavelet transform provides information on when the frequency change occurs, namely at time t1, when the value W1 jumps to the higher value W2. This time t1 of the transition from one wavelet value W1 to the other wavelet value W2 is utilized in the time-aligned arrangements of the determined power values for different frequencies described below. Figure 2 shows the processing device 5 in greater detail. A first processing subdevice 5a of the processing device 5 is frequency-based, while a second processing sub-device 5b of the processing device 5 is time-based. The first processing subunit 5a processes data from a frequency-based A / D converter module 8 of the sampling device 4, which samples the measured grid current and grid voltage L1, L2, L3, N, and ground—4-phase—in parallel at predetermined frequency intervals A, B, C, D, E, and F. The frequency ranges are divided as follows: Frequency Range 0: Direct Current (DC) Frequency range A: Sampling rate > 2.4 kHz in the range of approximately 0 to 55 Hz, or approximately 0 to 66 Hz—depending on the nominal frequency Frequency range B: Sampling rate > 24 kHz in the range of 47 Hz < 50 Hz < 53 Hz, or 55 Hz < 60 Hz < 66 Hz Frequency range C, D: Sampling rate > 40 kHz—typically 40.96 kHz in the range of nominal frequencies from 50 Hz to 2.5 kHz, or from 60 Hz to 3.0 kHz, in range D at intervals of 200 Hz Frequency range E: Sampling rate > 400 kHz—typically: 409.6 kHz, in the range from 9 kHz to 150 kHz Frequency range F: Sampling rate > 1.5 MHz, in the range from 150 kHz to 500 kHz Preferably, the frequency intervals within the individual ranges are equidistant, for example with an interval width of 1 Hz in frequency range A. The lowest frequency interval, for example, can range from 0.005 Hz to 1 Hz. The next frequency interval, one level higher, ranges from 1 Hz to 2 Hz, and so on, up to, for example, frequencies in the kHz range. Additionally, specific frequency ranges can be defined if particularly interesting frequency components are to be measured with very high accuracy. Figure 6 shows an overview of the various frequency ranges 0, A, B, C, D, E, and F. As described above, the sampling rate in frequency range B is higher than in frequency range A. Within each of the frequency ranges A, B, C, D, and E, the sampling rate is preferably constant, so that corresponding wavelet values can be determined for time intervals of equal length. Figure 7a shows the result of a wavelet analysis of a disturbance in the power grid performed according to Figure 2. In addition to frequency components—including a relatively high-frequency component in the operating band B between 47.5 Hz and 52.5 Hz—low-frequency components in particular were identified with frequencies of 25 Hz and below, which lie within the first frequency band A between 0 Hz and 50 Hz. The magnitude and distribution of the frequency components—that is, the frequency pattern determined on the basis of the wavelet transform—allow conclusions to be drawn about the nature of the fault. Furthermore, after a minimum latency period, the power at the start of the latency period can be determined. A closer look at the frequency components in the low-frequency range is shown in Figure 7b. It can be seen that a low-frequency resonance with sidebands is present, with a resonance frequency of 25 Hz. The sidebands are at 12.5 Hz and 37.5 Hz, as indicated by the hatched bar. The detection of such a low-frequency resonance with sidebands may be caused by modulation from electric vehicles connected to the power grid and other power-electronic loads. The first signal processing unit 5a also includes a first wavelet module 9. For each frequency interval, the digitized current and voltage values are multiplied by a predetermined wavelet function and then passed, in the order of the signal, to a first wavelet time-domain correction module 10. Figure 4 shows, by way of example, wavelet functions WL0, WL5a, and WL5b, which have different signal widths but the same signal height at time t0, when the amplitude is at its maximum. For example, these may be a Morlet wavelet or a Daubechies wavelet, which are particularly suitable for measurements of frequency intervals. The wavelet function can be scaled according to requirements, specifically with regard to signal width and signal level. The signal width depends on the frequency and is adapted to the frequency ranges so that the required accuracy can be achieved. The frequency-based first runtime correction module 10 is designed such that, depending on a frequency or frequency interval size, the determined wavelet value is shifted backward by a latency time 6a, 6b, 6c, 6d, 6e, 6f, so that the current and voltage values for each frequency interval are time-synchronized. The first processing sub-device 5a further comprises a processing module 11, by means of which the wavelet values representing the voltage and current values present for the respective frequency intervals are summed for the respective real and imaginary parts and calculated to yield a power value (active and reactive components). To determine the respective imaginary component, the wavelet values are each shifted by 90°. Thus, first wavelet values for a real component and second wavelet values for an imaginary component of the current and voltage can be determined. The specified latency time depends on the minimum frequency, which is determined by the lowest frequency interval. This is because, according to Heisenberg’s uncertainty principle, the lower the frequency, the longer one must wait before the complete measurement signal can be correctly recorded. Figure 3 shows, above the x-axis, the function of frequency versus the duration required to determine the current and voltage values in the respective frequency intervals. For example, a 0.1 Hz signal requires a duration of 10 seconds. A 0.01 Hz signal requires a duration of 100 seconds. These measured quantities in the frequency domain are further processed in processing module 11 to determine the power data. The second processing subunit 5b comprises a second wavelet module 12, by means of which the digitized current and voltage data provided by the A / D converter module 8 in the corresponding frequency intervals A, B, C, D, E, F into wavelet values At, Bt, Ct, Dt, Et, Ft using time-based wavelet functions. These wavelet functions are designed such that measured values can be determined relatively quickly at the expense of accuracy. A second runtime correction module 13, located downstream in the signal path of the second processing sub-device 5b, performs runtime correction so that the time- correct values for the respective frequency intervals can be superimposed or summed. At the output of the first runtime correction module 10 or the second runtime correction module 13, the resolved frequency components A, B, C, D, E, and F are summed to yield power and energy values for the respective intervals, which can then be used to determine the total energy or power. For this purpose, the processing module 11 includes a calculation module 15. The calculation module 15 includes a corresponding power calculation routine. Figure 8 shows an example of calculated wavelet values A1, A2, A3 at different frequencies, namely 0.3 Hz, 29 Hz, and 35 Hz. To determine the correct power, the latency time tl3 must be awaited for the lowest frequency, namely 0.3 Hz, and then the wavelet values A1, A2, and A3 must be summed for all frequency ranges. The latency time for the 35 Hz component, Tl1, is shorter than that for the 29 Hz component, Tl2. The longest latency time in this embodiment applies to the 0.3 Hz frequency component, Tl3. The latencies may, for example, correspond to the fixed sampling rates in the respective frequency bands A, B, C, D, E, and F. Preferably, the amplitudes A1, A2, A3 ... of the frequency components are compared with a predetermined threshold value S. If the determined amplitudes are less than the predetermined threshold value S, they are not taken into account in the calculation. In this embodiment, the amplitude A1 is less than the threshold value, so that the 35 Hz component is not taken into account in the calculation of power. Only the amplitudes A2 and A3 are taken into account for power determination. This threshold comparison allows insignificant frequency components—which might otherwise prolong the measurement duration—to be disregarded, since they are irrelevant for power determination. Figure 8 shows, as an example, the waveforms of three different wavelets, WL1, WL2, and WL3, during a fault condition. All three wavelet responses lie above a specified threshold value S. Wavelets lying below the threshold value S would not be included in the calculation. The first wavelet, WL1, corresponds to a measurement at a frequency of 35 Hz. The second wavelet, WL2, corresponds to a measurement at 29 Hz, and the third wavelet, WL3, corresponds to a measurement at a frequency of 0.3 Hz. Accordingly, this results in corresponding latency times tl1, tl2, and tl3, whereby, for the power measurement, the latency time tl3 of the lowest frequency must be waited out in order to obtain the power result at time t0 by summing the measured values determined for the individual frequencies. In addition, a database (not shown) may be provided, containing a trainable AI module designed to assign corresponding wavelet pattern functions to characteristic fault scenarios, such that the assigned wavelet functions are selected based on a recorded parameter characterizing a given fault. The AI module can be trained using an algorithm that processes a large number of training data sets, each of which contains training input values for input variables and training output values for output variables. Using the AI module, one or more input variables can be mapped to one or more output variables based on the parameters. It should be noted that the frequency-component current data and the frequencycomponent voltage data are each determined by wavelet function values at time t0, as shown in Figure 4. Time t0 is always the time at which the wavelet has its maximum amplitude. The length of a wavelet pattern is determined by the sampling rate, as shown in Figure 4. If the sampling rate is relatively high, as in frequency ranges C and D, the signal duration of the wavelet is relatively short. However, if the sampling rate is relatively low, as in frequency range A, the signal duration of the wavelet is relatively long; that is, it is temporally stretched compared to the wavelet with a high sampling rate. The invention utilizes a varying time and frequency resolution of wavelets to represent non-stationary signals, whereby the observation period and frequency are adapted to the signal being observed. The wavelet pattern functions stored in the database correspond to specific malfunctions. This is because if a match with a predefined wavelet pattern function is determined by comparing the currently derived wavelet pattern or frequency spectrum of the wavelet function, a characteristic malfunction can thus be easily detected. Such a fault can result from errors or changes in dynamics due to aging processes, the detection of defects in machine parts, or changes in the operating behavior of power electronic systems and devices. A prerequisite for this is that the wavelet pattern functions for these errors or faults are available or have been determined empirically. For example, these wavelet pattern functions can be empirically specified using corresponding measurement data or learned by the AI module. According to a further embodiment of the invention, a Daubechies wavelet is used as the wavelet pattern; this wavelet is designed as an orthogonal wavelet with a scaling function normalized to 1 and a wavelet function normalized to 1. The Daubechies wavelet WLD10 shown in Figure 9 is characterized by the fact that, as the frequency increases, the magnitude of the wavelet rises to the value +1 or, as shown in Figure 9, -1 at the desired frequency and then continues to increase the frequency. By comparing the respective current and voltage data at different sampling frequencies and applying the Daubechies wavelet to the corresponding sampled current and voltage values, the corresponding current and voltage values for the respective frequencies are, figuratively speaking, quantized. Since the Daubechies wavelet is normalized to a magnitude of 1 with respect to both the scaling function and the wavelet function, a root-mean-square value can be calculated for both the voltage V and the current I without the need for frequencydependent conversion. The following equations show the calculation of the root mean square value for the voltage V and for the current I, where the coefficients c and d correspond to the coefficients of the scaling function and the wavelet function of the Daubechies wavelet WLD10. The Daubechies wavelet WLD10 is known as wavelet D20 and has a total of 20 coefficients. %Ef = 100 J E Percentage energy of the wavelet coefficients for each level j frequency band Ej = j Cy(£)d£ and Ej = ^ c;2 (n) jR ££1 „ Fc-Fs i: — F Sampling frequency Root mean square value calculation based on Daubechies wavelet: V = 1 / 777 = 1^^+1^^ = 1^0 + 2^ / N 0 k k J j>jo = Jf^= J?^* +^= +l‘f k j>jo dj ,k = {v(t),(Pjk) dj,k = {i(t),<Pj,k} *iok^k Figure 10 shows an example of a Wavelet WLD10 with a sample frequency fS = 7.680 Hz. Here, different pseudo-frequencies and pseudo-time periods result for the various wavelet levels d1, d2, d3, d4, and d5. The smaller the pseudo-frequency—for example, at d5 = 164.2 Hz—the larger the pseudo-time period 0.0061 sec. Advantageously, the Daubechies wavelet pattern thus allows for the direct determination of the root mean square (RMS) values of current and voltage. Alternatively, it is also possible to determine the component of the current I and the voltage V as a function of specified frequencies. If, for example, the power for a specific frequency component is to be determined, only the corresponding frequency component of the current I and the voltage V is processed or calculated using the Daubechies wavelet pattern. In this process, a suitable sampling frequency is used with the A / D converter. Alternatively or additionally, a frequency-selective network impedance can be determined from the measured frequency-selective currents and frequency-selective voltages by calculating the corresponding ratio of the frequency-selective voltage to the frequency-selective current. In particular, the invention advantageously enables the determination of voltage and current values in the low-frequency range, i.e., preferably in a range between 0 Hz and 66 Hz, for example 0 Hz to 20 Hz. Due to the property of wavelets that they have a limited time duration in the range from 180 ms to 120 s (frequency interval 0 Hz to 66.0 Hz), the voltage and current values can be determined relatively quickly in all relevant frequency bands. The following examples illustrate different frequency ranges with varying wavelet time ranges: Area Frequency Interval Wavelet-Time Domain A 0Hzsfs55,0Hz 220 ms s T s 120 s A (US) 0Hzsfs66,0Hz 180 mssTs 120s B 47,0 Hz sU 53,0 Hz 226mssTs250 ms B(US) 55,0 HzsU 66,0 Hz 180 mssTs 220ms C 50 HzsU 2,5kHz 4,8 mssTs 240 ms D 2,5 kHz s U 9,0 Hz 1,3 mssTs 4,8 ms E 9,0 kHz s fs 150,0 Hz 0,08 mssTs 1,3 ms F 150,0 kHz sU 500 kHz 0,024 mssTs 0,08 ms It is evident that for the lowest-frequency range A, the wavelet has the largest time range and the longest duration. When using the Daubechies wavelet (DB10 or D20), the rms values of the measured voltages and currents can be determined directly, so that the rms value of the power can be calculated directly by multiplying the voltage and current values. As can be seen in Figure 9, the Daubechies wavelet pattern exhibits approximately 12 oscillations at the respective frequency; however, 8 to 10 oscillations can also be considered without significant loss of accuracy. A computer-readable medium—that is, a program (software program)—may be provided on which program segments executable by a computer unit are stored to carry out all steps of the described method for measuring power and / or energy when the program segments are executed by the computer unit. The program can be implemented in the device.
Claims
1. A device for determining power in electrical power grids (1) and / or forgenerating power patterns, comprising- a current measuring device (2) for measuring a grid current present at a measurement point of the power grid (1),- a voltage measuring device (3) for measuring a grid voltage present at the measurement point of the power grid (1),- a frequency-based A / D converter module (8) configured such that the grid current and / or the grid voltage are sampled at predetermined frequency intervals (Af) to produce digital frequency-component current data and frequency-component voltage data,- a wavelet module (9, 12) configured such that the frequency-component current data and frequency-component voltage data are multiplied by a predetermined wavelet function to produce wavelet frequency-component current data and wavelet frequency-component voltage data,characterized in that the A / D converter module (8) can be controlled such that it samples the grid current and the grid voltage in predetermined different frequency ranges (A, B, C, D, E) at different sampling rates to generate current data and voltage data, such that in a low-frequency range (A) the sampling rates are lower than in a comparatively higher-frequency range (B, C, D, E), and that a calculation module (15) is configured such that the wavelet-frequency-component current data and the wavelet-frequency-component voltage data, which have been multiplied in the wavelet module (9, 10), are summed to yield power data and / or energy data.
2. A device according to claim 1, characterized in that the temporal length of thewavelet pattern is selected such that it decreases continuously or discontinuously asthe frequency of the wavelet frequency component current data and the wavelet frequency component voltage data increases.
3. A device according to claim 1 or 2, characterized in that the time duration ofthe wavelet pattern lies in a range from 180 ms to 120 s for a frequency range between 0 Hz and 66 Hz.
4. A device according to any one of claims 1 through 3, characterized in that thewavelet pattern is configured as a Daubechies wavelet pattern, in particular as an orthogonal Daubechies wavelet pattern.
5. A device according to claim 4, characterized in that the Daubechies waveletpattern is configured such that both a scaling function and a wavelet function of the Daubechies wavelet pattern are normalized to a magnitude of 1.
6. A device according to any one of claims 1 through 5, characterized in that aruntime correction module (10) is configured to shift the frequency-component current data and the frequency-component voltage data by a latency time as a function of a frequency (f), such that the frequency-component current data and the frequencycomponent voltage data of the frequency intervals are available to the calculation module (15) in a time-synchronized manner.
7. A device according to any one of claims 1 through 6, characterized in that theA / D converter module (8) is configured such that the grid current and the grid voltage are sampled at equidistant frequency intervals (Af).
8. A device according to any one of claims 1 through 7, characterized in that thewavelet module (9) is configured such that the predetermined wavelet function is shifted by 90° in the frequency domain to provide a first wavelet function for a real part and a second wavelet function for an imaginary part of the current or voltage data, respectively.
9. A device according to any one of claims 1 through 8, characterized in that thepredetermined latency time depends on the minimum frequency of the frequencycomponent current data or frequency-component voltage data, respectively.
10. A device according to any one of claims 1 through 9, characterized in that thelatency time is determined based on the Heisenberg uncertainty principle.
11. A device according to any one of claims 1 through 10, characterized in that asecond wavelet module (12) is provided such that the frequency-component current data and the frequency-component voltage data are shifted by a latency time as a function of a time value, so that the frequency-component current data and frequency-component voltage data for the respective frequency intervals are available at a predetermined time within a time interval following the occurrence of a fault in the grid, wherein the length of the time interval is determined by the smallest frequency component of the current signal or voltage signal to be monitored or evaluated, respectively.
12. A device according to any one of claims 1 through 11, characterized in thatthe wavelet function is weighted and / or validated as a function of the frequency components.
13. A device according to any one of claims 1 through 12, characterized in thatonly current data and voltage data for frequency components that lie above a predetermined threshold value (S) are taken into account in the calculation and / or evaluation.
14. A device according to any one of claims 1 through 13, characterized in that adatabase is provided in which a trainable AI module is arranged, which is designed to assign corresponding wavelet pattern functions to characteristic fault constellations, such that the assigned wavelet function is selected as a function of a detected parameter characterizing a predetermined fault.
15. A device according to any one of claims 1 through 14, characterized in thatthe AI module is trainable by means of an algorithm that processes a plurality of training data sets, each of which contains training input values for input variables and training output values for output variables, and which is configured to map one or more input variables to one or more output variables based on the parameters.
16. A method for performing a power measurement in power grids, wherein gridcurrent values and grid voltage values are sampled and digitized, characterized in that- the grid current values and the grid voltage values are sampled in predetermined frequency ranges (A, B, C, D, E) at different sampling rates, wherein in a first frequency range (A), in which relatively low-frequency components of the grid current values and the grid voltage values are present, the sampling rate is lower than in a second frequency range (B, C, D, E), in which the frequency components of the grid current values and the grid voltage values are comparatively larger,- that the sampled grid current values and the sampled grid voltage values are each multiplied by a wavelet pattern to produce wavelet current data and wavelet voltage data, which each form a wavelet curve (WL1, WL2, WL3),- that the wavelet waveforms (WL1, WL2, WL3) are synchronized with one another, wherein the wavelet waveform (WL3) with the lowest sampling rate serves as a reference for the other wavelet waveforms (WL1, WL2) with comparatively higher sampling rates,- that the power is calculated by summing the wavelet current data and wavelet voltage data characterizing the respective wavelet profiles (WL1, WL2, WL3).
17. The method according to claim 16, characterized in that the grid currentvalues and the grid voltage values are multiplied by an orthogonal Daubechies wavelet pattern, whose scaling function and wavelet function are normalized to 1 in magnitude, so that effective values of the current data and the voltage data can be directly determined therefrom.
18. The method according to claim 16, characterized in that the same waveletfunction is multiplied by the measured grid current data and grid voltage data at time intervals that are equidistant at least within the frequency ranges.
19. A method according to any one of claims 16 through 18, characterized in thatonly those wavelet profiles (WL1, WL2, WL3) of grid current values and / or grid voltage values that lie above a predetermined threshold value (S) are taken into account in the power calculation.
20. A method according to any one of claims 16 through 19, characterizedin that wavelet curves (WL1, WL2, WL3) in a frequency range comprising lowerfrequencies are taken into account in the calculation with a longer latency time (tl3) than the wavelet curve determined at a higher frequency.
21. A computer-readable medium on which program segments executable by acomputer unit are stored to perform all steps of the method according to any one of claims 16 through 20 when the program segments are executed by the computer unit.