RU2339046C2 - Electrical energy quality monitoring - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: data (step 202) representing controlled signal cycles are collected, electrical energy signal waveform is expanded (successive steps 204, 206, 208, 210, 212, 214) to set of components. Thus this set includes the whole spectrum of wave cycles. Further these values are individually compressed (step 216), and these values and related variation ranges values are saved (step 218).

EFFECT: higher efficiency of waveform value compression of traced signal of electrical energy.

49 cl, 4 dwg

Description

FIELD OF THE INVENTION

The present invention relates to the field of monitoring the quality of electricity, in particular to storing the data obtained for subsequent use.

State of the art

To ensure the high quality of electricity supply services, it is desirable to monitor the quality of electrical signals (i.e., electrical energy signals) at various points in the electrical network. In particular, it is desirable to know the waveform of the current and voltage at various points in the network during the occurrence of any specific anomaly (for example, a power outage) in order to identify its cause.

US Pat. No. 5,736,647, incorporated herein by reference, describes a measuring device for reading and displaying the shape of electrical energy signals, and for calculating their parameters. These parameters may include, in particular, current, voltage, power, including in reactive kilovolt-amperes (kVAR), multiphase signal phase ratio, power factor, frequency, electric energy in kilowatt / hour (kW / h) and kilovolt-amperes / hour (kVA / h) and reactive energy (in kVAR / h).

However, in some cases, real-time display of the waveform and / or derived parameters is not enough. US Pat. No. 6,657,071, incorporated herein by reference, states that it would be desirable to use an electric power quality control device that constantly monitors all voltage waveform data. However, due to the large volumes of processed data, such a device is practically impracticable. At a sampling rate of 32 measurements per cycle per second, 1920 measurements are taken. For three-phase measuring instruments measuring waveforms of both voltage and current, such data is 6-8 times more. This patent proposes the detection of waveforms and the collection of data relating to time intervals corresponding to the detection of changes.

US Pat. No. 6,671,654, incorporated herein by reference, describes an electronic device that is designed to measure the parameters of electricity coming from a power distribution system. As in the previous patent, the data is saved only for the anomalies that have occurred (for example, for an abrupt change in current or voltage).

US Pat. No. 6,639,518, incorporated herein by reference, describes a network monitor that monitors one or more circuit parameters. The specified monitor contains a system for selecting the moments of the alarm.

Some monitoring systems store only rms (SK) values and / or power of electrical signals. However, to accurately identify the nature of the problems, this data is not always sufficient.

Disclosure of invention

One aspect of some embodiments of the invention relates to a device for continuously memorizing the waveform of one or more electrical signals (electrical energy signals), regardless of whether an anomaly of interest is identified. If necessary, the device may contain memory, the amount of which is sufficient to store the waveform of the signal corresponding to the normal modes of power supply for a long period of time (at least a month, two months or even a year). The storage of the indicated data, covering such a long period, allows one to return to the anomalies that have occurred, to determine various parameters that are considered important during the anomaly, and / or to reanalyze the anomalies, which at that time were not considered significant or were not detected at all.

In some embodiments of the invention, the data of continuously stored signals are compressed (compressed) using an appropriate method, taking into account the nature of the electrical energy signals, if necessary. It is possible to implement a real-time compression method as signals are collected, i.e. compression operations are carried out until all compressed data is received. For any data segment, compression can be performed on the basis of a small number of them. In some embodiments, the amount of uncompressed data used by the processor that compresses, generates, in one form or another, the compressed data is relatively small, covering for uncompressed data, for example, a period of less than a minute, a second, or even 0.1 s. If necessary, the amount of memory that is acceptable for use and / or actually used for compression by an appropriate processor is less than 100 kB or even less than 10 kB.

According to some embodiments of the invention, the computing capabilities and / or engineering resources that have been spent so far on identifying anomalies are sent to compress and store large amounts of information related to storing waveform data (waveform).

The waveform can be stored with high accuracy by conducting at least 64, 128, 256, or even 512 measurements during the electric cycle (usually 1/50 s or 1/60 s).

One aspect of some embodiments of the invention relates to the use of a lossy compression method for compressing traceable electrical energy signals. This application of this method allows for a high compression ratio, i.e. storage of large amounts of data with a relatively small decrease in quality. In some embodiments, instead of accurately preserving a limited number of parameters, a waveform of electrical energy is retained while allowing a slight loss in accuracy. If necessary, if more accurate information is also required, the exact values of one or more of the most important parameters are stored together with the waveform. Upon reaching a compression ratio of 1: 500, 1: 1000 or higher, a decrease in accuracy due to lossy compression can be up to about 1-2%, preferably up to about 0.1%.

As one option, lossy compression is performed in real time. In some cases, this may require a compression unit that meets the requirements of relatively high processing power.

One aspect of some embodiments of the invention relates to real-time compression of monitored electrical energy signals. Such compression is carried out promptly (during operation), and it is not required that the compression unit has a large internal memory. In the alternative or additional order, real-time compression is carried out for each measurement within the interval of less than a second or even half a second, counting from the moment it is received.

One aspect of some embodiments of the invention relates to the compression of electrical energy signals by decomposing them into many components (for example, components that differ in frequency) and separately compressing each component over a relatively long time span. In some embodiments, compression converts the values of each component over time into a model that can be generated using a small amount of data. For example, components that have the same values can be combined by fixing these values and the lengths of time for which they appear. Compression of the values of each component can be carried out without correlating them with the values of any other component. As one of the options, the signal of electrical energy in each cycle is decomposed into components and compresses the values of the components covering many cycles.

If necessary, compression is lossy compression that ignores small errors when fitting component values to models. In some embodiments, such compression ignores errors that are less than the accuracy of the instrument measurements.

In some embodiments, said plurality of components are frequency components (i.e., harmonics). Such components can be obtained by applying a transformation such as the Fourier transform (e.g., DFT, discrete Fourier transform, or FFT, fast Fourier transform), and the cosine transform (DCT, discrete cosine transform) or Z-transform to measurements in the time domain. Alternatively or additionally, frequency components are obtained using a set of filters (discrete or analog) corresponding to different frequencies. In addition, in the order of an alternative or additionally, these components are obtained using the procedure of pattern recognition in time. An alternative for frequency components is time-domain components, such as fragments of a cycle of an electrical energy signal.

As one of the options, the signal of electric energy is characterized by the value obtained by storing, for example, the average values for each harmonic, and the corresponding amounts of successive cycles in which the harmonic has a value close to the average value. Although the Fourier transform is a relatively intensive computational process, the obtained compression ratio and, thus, the ability to store electrical energy signals spanning large periods of time are considered quite acceptable.

In some embodiments of the invention, the cycle time of the electrical energy signal (i.e., the proximity of the cycle to a nominal value of 50 or 60 Hz) is monitored continuously. As a result, the Fourier transform is carried out on the cycle signals in a strict manner, without distorting the cycle. In a number of embodiments, the frequency of measurements of electrical energy signals is adjusted according to the length of the cycle, so that each cycle, regardless of its length, corresponds to essentially the same number of measurements. If necessary, each cycle corresponds to a predetermined number of measurements acceptable for a fast Fourier transform (for example, 256 measurements).

As a result of the exact correspondence to the cycles of electric energy signals, for long intervals of the signals, the harmonics usually have the same or very close values. Thus, if only changes are saved, compression with very high coefficients (for example, 1: 10000) is achieved for normal signals of electric energy, which allows storing large amounts of data.

In some embodiments, harmonics having very low (i.e., close to zero) values are not retained. Typically, many harmonics of the cycles of electrical energy signals have substantially zero values. Therefore, they do not require space in the amount of memory, especially when the controlled electrical lines function satisfactorily.

One aspect of some embodiments of the invention relates to storing a waveform of an electrical energy signal for a long period of time (at least a week or even a month) by storing the spectral components of the signal. It is possible to compress said components.

Thus, according to one embodiment of the invention, there is provided a method for compressing values characterizing the waveforms of a monitored electrical energy signal, comprising: collecting data representing periods of the electric energy signal, decomposing the waveform into a plurality of components spanning the plurality of waveform periods, and separately compressing the values at least some components spanning many periods.

As one of the alternatives, the decomposition of the waveform of an electrical signal into components is a decomposition into frequency components. If necessary, compressing the values of at least some of the components includes fitting the time segments of the components to the model and registering the coefficients of such fitting. A model may be a constant time function. As will be explained below, the registered coefficients for such a function may be a single value and length. As one of the options, the model is a monotonic function of time. The collected data, representing the periods of the signal, may contain recorded measurements of the electrical signal, and the measurements are divided into groups corresponding to the cycles of such a signal. The decomposition of the waveform can be a transformation of the measurements of each group into the values of the harmonic components.

As one of the options, for each harmonic, separate compression of at least some components includes the preservation of pairs, each of which consists of an average value and the number of cycles in which the tracked value is close to the average value. The number of such cycles can be established by evaluating the minimum and maximum sequences of harmonic values and determining the moment when the difference (distance) between the minimum and maximum exceeds a given level. The indicated difference can be expressed as a percentage of the previous average harmonic value or of the expected harmonic value.

A variant is possible in which the collection of discrete measurements (sampling) is the sampling of analog signals. The division of measurements into groups can be a repeated determination of the main network frequency of the signal and the corresponding determination of the signal cycles of electric energy.

In some embodiments, the signal sampling is carried out with a frequency determined in accordance with the frequency of the network. Repeated determination of the main frequency of the network can be made on the basis of the received measurements. As an alternative or additionally repeated determination of the main frequency of the network can be made on the basis of an analog signal to which the received measurements correspond. Among other options, the measurement transformation of each group can be a transformation that uses the fast Fourier transform.

As one alternative, the method includes applying a lossless compression procedure to the compressed harmonics. For this method, it is possible to store at least some of the compressed components in a file structure representing a plurality of electrical signals. On the other hand, the method may include storing the compressed components in a file structure representing a signal of electrical energy continuously for a time interval exceeding a month. A possible variant of the method with the transmission of compressed components over the communication line. At least for some components, compression can be performed separately for each component. In another embodiment, at least some components are compressed in real time and / or using lossy compression. It is possible to compress at least some components separately from each other with overlapping at least three waveform periods. Alternatively, the electrical energy signal comprises a current signal or a voltage signal. The collected data representing the periods of the signal may include collected data representing cycles of the waveform or, alternatively, periods that are shorter or longer than the cycles of the wave.

A variant is possible in which the decomposition of the waveform into many components is carried out in the form of decomposition into components coexisting in time.

According to one embodiment of the invention, there is further provided a method of compressing values of a monitored electrical energy signal, including collecting measurements (sampling) of an electrical energy signal, dividing the measurements into groups corresponding to the cycles of such a signal, converting the measurements of each group into harmonic component values, and storing the display of harmonic component values in non-volatile storage medium continuously for at least a week.

As one of the options, storing the indicated display is storing a compressed display of the values of harmonic components. Memorizing a compressed display may be a memorization of compression based on the joint compression of the values of each harmonic component spanning multiple cycles.

Further, according to one embodiment of the invention, there is provided a device for monitoring electrical energy signals. The device according to the invention comprises an electric line interface for measuring electric line signals, a non-volatile storage medium and a processor configured to record information on the waveform of the measured electric line signals continuously into the non-volatile storage medium, regardless of whether an anomaly has been identified.

As one of the options, the electric line signals are at least one current signal and / or at least one voltage signal. An embodiment of the processor is provided with the ability to record the specified display continuously for at least a week. The device may not be adapted to detect anomalies. In one embodiment, the storage medium has a capacity of less than 1 GB, and in the alternative, more than 1 GB. The processor may be configured to compress the measured electrical energy signals with a predetermined at least minimum compression ratio.

A variant is possible in which, in order to achieve a predetermined minimum compression ratio, the processor is adapted to control the level of compression loss. The processor may also be configured to compress measured electrical energy signals in real time or using a lossy compression method. It is possible to perform an interface with the ability to receive measurements of electrical energy signals, at least with a frequency of 8 measurements per cycle of these signals.

In addition, according to one embodiment of the invention, there is provided a method of compressing values of a monitored electrical energy signal, comprising collecting measurements of an electrical energy signal and compressing them using a lossy compression method. Measurements can be compressed in real time. In one embodiment, the collection of measurements is carried out with a frequency of at least 50 measurements per cycle of the electrical energy signal.

Brief Description of the Drawings

Next will be described examples of embodiments of the invention, which should be considered in conjunction with the accompanying drawings. Identical structures, elements, or parts found in several drawings have the same or similar numerical designations in all of these drawings.

Figure 1 is a schematic diagram of a compression unit according to one embodiment of the invention.

Figure 2 is a block diagram of a compression method according to one embodiment of the invention.

FIG. 3 is a flowchart of a method for determining time intervals corresponding to one embodiment of the invention during which harmonic or parameter values are substantially the same.

Figure 4 schematically illustrates the structure of stored compressed data according to one embodiment of the invention.

The implementation of the invention

1 schematically shows a compression unit 100 according to one embodiment of the invention. It is possible to connect the indicated unit to electric wires, for example, to three wires 102 corresponding to different phases, and to a neutral line 104, in order to monitor the signals passing through the wires in order to monitor the quality of electric power. The compression unit 100 compresses the signals received from the wires, thereby allowing the storage of data monitored for extended periods of time. At the same time, it is possible to continuously store data regardless of whether an anomaly of interest has been identified. In some embodiments, the compression unit 100 comprises connections 130 and 132 for measuring current and voltage, respectively. These compounds can be material in nature, carried out by means of a magnetic field and / or represent any other modification of the bond.

A variant is possible in which the compression unit 100 comprises an analog sensor 110, which is sensitive to monitored signals, an ADC (analog-to-digital converter) that converts the sampled analog signals to digital measurements (samples), and a compressor 112 that performs compression. In some embodiments, the compression unit 100 further comprises a processor 118 that determines one or more parameters of the raw recorded data. These parameters include power, phase ratio between the components of the multiphase signal, power factor, main frequency of the network, energy and power (in kW / h, kVA / h, kVAR / h, kW, kVA, kVAR), total non-linear distortion coefficient, rms (SK) current and SK voltage. For each phase, at least some of these parameters can be calculated. Alternatively or additionally, the list of monitored parameters contains factors characterizing the decline, increase and / or runout, and / or data used to determine these parameters. In some cases, the relationship between the values for different phases is calculated. The parameters calculated by processor 118 may be compressed by compressor 112 and / or used in a manner known to those skilled in the art to generate alarms. The parameters that are calculated using the processor 118 may be those having great utility, or parameters that are relatively largely affected by small waveform errors. If necessary, these parameters are saved in addition to the waveform itself. This avoids the errors caused by the lossy compression to which the waveforms are subjected.

Among other options, the compressor 112 performs real-time compression, i.e. signals are compressed in less than a second or even a tenth of a second, counting from the moment they are collected. The compressed data obtained from the compressor 112 may be stored in a storage unit 120 of any type known in the art, for example, a drive, a compact flash memory, or any other non-volatile memory. In some embodiments, the storage unit 120 has a relatively large capacity, i.e. able to store data collected for several months or even for a period exceeding a year. The capacity of such a unit can exceed 1 GB, 10 GB or even 40 GB. In alternative embodiments in which the compression by compressor 112 is very large, a relatively small storage unit 120 with a capacity even less than 1 GB is sufficient.

In some embodiments of the invention, the compressor 112 comprises a cache memory that stores data until it is recorded in the memory unit 120. Alternatively or additionally, the cache stores data necessary for performing compression. It can have a capacity of less than 10 MB, 1 MB, or even 100 KB.

In alternative embodiments, or in addition to storing the compressed signals, such signals are supplied to a transmitter 124 for transmission to a remote device. Before such a transmission, the signals are compressed to allow transmission at an acceptable bandwidth. Transmission can be accomplished using any communication channel known in the art, including controlled electrical lines, radio lines, cable and / or telephone lines. In the remote device, the transmitted signals can be stored, displayed and / or analyzed. In some embodiments of the invention, the compressed signals stored in the storage unit 120 are open for access from the control panel 126, which can be used to display information and / or to analyze the compressed signals. As will be shown below, the signals can be stored in such a way that allows you to quickly search for signals related to a specific date and / or having a given type. As one option, the remote control 126 is also used to adjust the performance of the compression unit 100.

The analog sensor 110 may be of any type known to those skilled in the art, including the sensors described in the aforementioned US Patent No. 5736847. Compressor 112, if necessary, has sufficient performance to perform Fourier transform (eg, FFT) of each of the monitored signals, and for each cycle of the electrical signal.

ADC 114, if necessary, can operate with a frequency sufficient to block the desired number of harmonics of the received signals. In some embodiments, the ADC 114 operates at a predetermined sufficiently high frequency. In alternative embodiments, the frequency of the ADC is regulated in accordance with the main network frequency of the measured signals. If desired, for different lines and / or different signals (for example, current, voltage) on the same line, you can use a different sampling frequency. Alternatively, the same sampling frequency is used for all signals measured by analog sensor 110.

ADC 114, essentially, can have any acceptable bit depth, for example 16 bits. As one of the options, this parameter is selected based on the accuracy of the measuring equipment.

Figure 2 is a block diagram of a compression method according to one embodiment of the invention. The main frequency of the network is determined at step 204 based on the signals measured at step 202, using any method known to specialists in this field, for example, any method based on phase-locked loop (PLL), and using characteristics such as detecting zero crossing if necessary level or autocorrelation. The sampling frequency (sampling frequency) corresponding to the main frequency of the network is calculated at step 206. If at step 208 it is found that the calculated frequency is significantly different from the actual sampling frequency of the ADC 114, the sampling frequency is changed at step 210 to the calculated frequency immediately after detecting that a change is required, or after one or more cycles. In an alternative embodiment, the change is carried out gradually, using, for example, a hysteresis algorithm.

At step 212, measurements of each cycle of the recorded signals are converted into the frequency domain (i.e., into harmonic vectors formed from real and imaginary values), using, among other options, fast Fourier transform (FFT). Alternatively or additionally, in step 214, the values of one or more parameters of the registered signals are determined. For each parameter and harmonic, the compressor 112 determines in step 216 the time interval (or the number of measurements) during which their value does not change significantly, tracking the values and the corresponding interval lengths of the equivalent values. Next, with reference to FIG. 3, one example of a variant for determining the lengths of such time intervals is described, namely, step 216. As one of the options, the signal values and durations of the time intervals determined at this stage are stored in step 218 in an acceptable file format, as this, for example, is described below with reference to figure 4. In some embodiments, the file format is further compressed using any suitable lossless and / or general purpose compression method (i.e., independent of the type of data represented by the compressed signals). Such methods, in particular compression according to the so-called Lempel Ziv Welch algorithm (Lempel-Ziv-Welch, LZV), are known to specialists in this field.

In a more detailed discussion of the calculation procedure at step 206, the sampling frequency (sampling frequency) when determining the fundamental frequency, it should be noted that in some embodiments of the invention, this frequency is calculated based on the number of harmonics used in compression. In one embodiment, 128 harmonics are used, although the number of harmonics can be increased to achieve more accurate compression and / or to capture high frequencies and / or transient states. For example, at a main frequency of 50 Hz, a sampling frequency of 1 MHz corresponding to approximately 20,000 harmonics can be used to capture fast transients. Since higher harmonics usually have zero values, the size of the compressed signals from the application of such a large number of harmonics is essentially independent. In the case where high accuracy is not required, as an alternative, fewer harmonics (e.g., 32) can be applied to lower the performance required by compressor 112 for performing Fourier transforms. As one option, the sampling frequency used is equivalent to the product of twice the main frequency of the network by the number of harmonics, so for each cycle of the monitored signals there is a number of measurements sufficient to construct the desired number of harmonics. For example, for electric lines operating at frequencies of 50 Hz and 60 Hz, to obtain 128 harmonics, sampling frequencies of 12.8 kHz and 15.36 kHz, respectively, are used.

Upon a more detailed consideration of the procedure for detecting the fact of a significant difference between the calculated and real sampling frequencies (step 208), it should be noted that in some embodiments of the invention, the values are considered significantly different, based on the effect of this difference on compression. Among other options, when this difference causes a high probability that subsequent FFT cycles will have different values, even if the cycles of the monitored electrical energy signals are essentially the same, the sampling frequency by ADC 114 is changed at step 210.

In addition to determining the main frequency of the network based on the measured signals, as an alternative, the indicated frequency is found based on analog signals using methods known from the prior art. As one of the options, in the case when the compression unit 100 operates with a plurality of electric lines, the sampling frequency is determined based on the average fundamental frequency of the lines or on the basis of the fundamental frequency of one of the lines used as representative, since the frequencies of the connected lines are usually are the same. Alternatively or additionally, for example, if the various electric lines are not connected, each of them uses its own sampling frequency.

An alternative to adjusting the sampling frequency in accordance with changes in the fundamental frequency is a situation where a sampling frequency that is higher than the required one is used and the fundamental frequency is used to group the measurements into cycles. According to such an alternative, instead of the FFT, a discrete Fourier transform (DFT) or some other type of transformation, for example, Z-transform or DCT, can be used. As an alternative, measurements are repeated so that their number meets the requirements of the FFT.

In some embodiments of the invention, when the synchronization with the signal is lost, on the basis of which the main network frequency is determined, the last measured network frequency is used in the cycle length determination procedure (for other monitored signals) until synchronization is restored.

An alternative to using FFT is the possibility of using any other method that can be used to transform the values of each cycle into harmonics. These alternative methods can be, for example, other transformations, in particular the Z-transform or DCT. Alternatively or additionally, filter sets and / or element recognition over time are used for the indicated transformation. When examining in more detail the procedure for determining (step 216) the lengths of time intervals within which the harmonics do not change significantly, it should be noted that in some embodiments of the invention, when the values vary slightly according to a predetermined pattern, this is fixed in a compressed format . As a result, signals can be more accurately reconstructed and / or tighter tolerance values applied.

Alternatively or in addition to the above procedure, at step 216, the compressor 112 may attempt to fit the measured values to a predetermined function or to a predetermined pattern, for example in the form of a monotonically rising or falling line and / or a low polynomial. In addition, as an alternative or additional way, the compressor 112 from the predefined set of functions selects the function that best suits the values. Compression results include an indication of the selected function and applied coefficients.

Figure 3 presents a flowchart of one of the variants of the method for determining (step 216) the lengths of time intervals during which the values of the harmonic or parameter (to simplify the presentation, the common joint term "harmonic" is used below) remain essentially the same. In the procedure illustrated in FIG. 3, the variables min, max, acc and avg correspond to the current minimum, maximum, total and average values of the harmonics in the time interval in which they are essentially the same (i.e., the difference between the minimum and maximum values is small). The variable count stores the number of samples in the current series.

After starting at step 300, when registration of a new signal begins (as described below with reference to FIG. 4) and when a substantially different value is identified for the harmonic (moment 330), the values of the new measured signal are assigned to the variables min, max and acc (step 302 ), to denote which is further used the variable new, i.e. min = max = acc = new. Then, the next new value (new) is recorded (step 303) and compared with min and max (step 304) to determine whether the value of any of these variables changes, i.e.

if new <min, then min = new,

if new> max, then max = new.

The difference between the minimum and maximum values is compared with the maximum allowable value of this difference (step 306), and these values are still considered to belong to the same series. If this difference exceeds the specified tolerance, i.e. | max-min |> tolerance, the previous series is closed (step 310) and a new series is started (step 302). The procedure for closing the previous series at step 310 includes storing the average value (avg) of the series of measurements along with the number of (count) samples. Alternatively, or in addition to maintaining the mean value, the total value (acc) is also stored. If the difference between the minimum and maximum values is within the specified tolerance (step 306), a new value is added to acc (step 315), and the parameter count is increased by 1. Then a different harmonic value is obtained (step 303).

After performing many cycles of the method presented in figure 3, get a set of pairs of values avg and count. Typically, in signals corresponding to electric power, the same or very similar harmonics appear for many consecutive cycles of the main frequency of the network. Therefore, the representation of many cycles of an electrical signal can be obtained using relatively small amounts of these pairs of values and the successive cycles in which they occur.

It should be noted that for some harmonics, the equivalence of values is especially pronounced, so even if for some harmonics the gain in the required memory due to the compression method of the present invention is small, for a number of other harmonics it is very significant.

In some embodiments, instead of preserving the harmonic values themselves, the relationship between the two harmonics is known, which is known to be interconnected. In some cases, this ratio is more stable compared to the values themselves. In one embodiment, the relationship between the higher harmonics and the fundamental is applied. Alternatively or additionally, the compression unit 100 decides in real time whether to keep the harmonic values themselves or their ratios depending on which option currently provides more stable values.

In some embodiments, the tolerance value (tolerance value) is set as a percentage of the nominal harmonic value. Alternatively, this value is specified as a percentage of the fundamental or of the total signal. The tolerance structure can be formed according to the desired compromise between accuracy and compression ratio. It is possible to take into account the measurement error of signals in this structure so that the tolerance (in percent) is not less than the specified error. It should be noted that in situations in which the error is relatively high (for example, 0.1% of the measured values), the tolerance value can be set at the level of the own error of the measuring equipment. Then lossy compression does not lead to loss of information. In one embodiment, a tolerance in the range of 0.1-1% of the nominal value is used. An alternative to using the generated tolerance structure is to adjust its value over time according to the compression parameters. So, in one embodiment, the tolerance value is adjusted in accordance with the obtained compression ratio. A variant is possible in which, in order to obtain the desired compression ratio at each completion of the data recording 414 (see FIG. 4), the compression ratio of said registration is determined and the tolerance for the next data recording is selected accordingly.

It is possible to use the same tolerance for all compressed signals. Alternatively, for different harmonics and / or parameters, different tolerance values are used, for example in accordance with the importance of each harmonic and / or each parameter. As a nominal value, a predetermined expected harmonic value is preferably adopted. As an alternative, the indicated value is the average value determined based on the harmonic value in the last cycles.

In some embodiments of the invention for harmonics having complex values, the method presented in FIG. 3 is implemented separately for the real and imaginary components of harmonics. In the alternative order, the method is implemented without separating the complex values of the harmonics and comparing with the tolerance based on the difference between min and max. In addition, as an alternative or additionally specified method is implemented separately for amplitudes and phases of harmonics.

4 schematically illustrates a compressed data storage system 400 according to one embodiment of the invention. As one of the options, compressed data can be placed in files (FILES) 404 corresponding to sessions (SESSIONS) 402. Files 404 corresponding to one session 402 are formed sequentially, i.e. there is no interval between the data of such files. However, between sessions 402 there may be an interval with missing data, for example, due to an instant shutdown of the system and / or a stop / start command issued by the user.

Each file 404 may contain a header (HEADER) 410, an information record (INFO) 412 identifying the types of data that are located in the file, and data records 414 in the form of blocks containing compressed data (DATA). In some embodiments, each data record 414 contains a header 420 and signal fields 422 (denoted 422A, 422B, 422C, etc.), each field 422 corresponding to one signal (SIGNAL) monitored by the compression unit 100. For signals related to a single significant parameter and illustrated, for example, by a 422 V signal field, the signal field contains a length field 430 (LEN) and an array (ARRAY) of 432 pairs of average values (avg) and the number of consecutive occurrences of the corresponding values in the tracked signal (count see Fig. 3). The values of the array 432 are stored in step 310 (FIG. 3). In some embodiments, the average values are fixed by indicating the ratio of the average value to the nominal value of the signal. Applying this ratio reduces the amount of memory required to store the average value.

Signal fields 422 related to waveforms represented by their harmonics are shown in FIG. 4 as field 422D, which contains many harmonic fields 438 (these fields correspond to signal harmonics) and a numeric field 434 indicating the number of fields 438 (NUM . HARMONICS) in field 422D. Among other options, each field 438 contains an identifying field (ID) 436 indicating the harmonic to which field 438 relates, field 432 of an array (ARRAY) of data containing pairs of average values and the number of times they appear consecutively, and also field 430 of length (LEN ) indicating the number of pairs in field 432. In some embodiments of the invention, harmonics having zero values are not input to signal field 422D. Usually, in the signals coming from the electric line, many higher harmonics are of zero value, therefore, the refusal to remember such harmonics is an advantage of the invention. Alternatively, the signal field 422D comprises a harmonic field 438 intended for each of the calculated harmonics. In this case, all 422D fields have the same number of harmonics, i.e. identification fields 436 and numeric field 434 are not required. In another alternative situation, if useful, a numeric field 434 indicates that there are values for all harmonics, and identifying fields are not applied.

In some embodiments, when the harmonic values change essentially in each cycle over a relatively long period (for example, in excess of 10 cycles), a different storage format is used, for example, storing only the measured values and / or storing the difference values between successive values. It is possible to use a special indication indicating that a different format has been applied to the signal.

In some embodiments of the invention, the order of harmonics within the 422D signal fields is arbitrarily selected, for example, starting from the fundamental and sequentially reaching the highest (for example, to the 127th harmonic). As an alternative, an order is selected that is expected to ensure that the file achieves improved lossless compression, for example, in the form of LZV compression. You can also select the order of harmonics according to the lossless compression method applied to file 404 based on the expected harmonics similarity. In one embodiment, the odd harmonics are stored separately from the even ones. Alternatively or additionally, harmonics are arranged in descending order.

In some embodiments of the invention, substantially all of the data records 414 contain data on the same signals. In one embodiment, the signals stored in the data records 414 comprise current current and voltage waveform signals represented by long signal fields 422D. In addition, the signals stored in the data records 414 may comprise rms values of voltage and current. Alternatively, various data records 414 relate to different groups of signals.

A variant is possible in which the header 420 of the data record 414 indicates the start and / or end time of the data arrival to which the specified record belongs. In some embodiments of the invention, in order to indicate the time in the header 420, a compression block circuit 100 using internal time is used. As will be shown later, the information record 412 of each file can record the local (i.e., reference) time and the corresponding internal time intended for use in converting the internal time to the standard (zone). The use of internal time within the boundaries of the headers 420 allows using a shortened time field and / or eliminating the possibility of a time difference caused by an erroneous setting of the zone time by the operator during data collection of a single file. Local time allows the comparison of compressed data with data from other compression units 100 and / or with data about other anomalies. As one option, local time is indicated in each file 404, so that misphasing of the clocks of the internal time does not cause a large drift of time indicated for the collected data.

Each of data records 414 of file 404 may relate to signals of a particular time period, i.e. all signals of a given period are placed in a single data record 414. Records 414 of a single file 404 can be arranged sequentially in accordance with the period of time to which they relate. As one option, various data records 414 relate to time periods having different lengths. In some embodiments of the invention, a restriction is introduced according to which data records 414 relate to data obtained over a period of time not exceeding a predetermined duration, in particular 2-10 minutes, for example 5 minutes. In some embodiments, a predetermined maximum time period is selected so as to limit the time interval for the fundamental frequency of the network to change, since arrays 432 use the number of cycles as an indicator of time.

Alternatively or additionally, data records 414 should reach a maximum amount lying, for example, in the range of about 0.5-2 MB. In another alternative or additional situation, the records 414 are limited by the volume of the original uncompressed data so that the data of a single data record 414 is not difficult to manipulate in one processor. In one embodiment of the invention, data records 414 are limited to an upper level in the range of 2-20 MB of data.

Among other options, files 404 are limited to an upper level not exceeding the volume, which in practice is determined by the capabilities of a computing unit (for example, which removes compression) and a memory unit. The specified level can reach, in particular, 20-40 MB. Alternatively or additionally, these files are limited by the period of time to which they relate (for example, no more than a day or a week) and / or by the number of data records 414 contained therein. In some embodiments, the implementation of these parameters on the files 404 do not impose restrictions.

In some embodiments of the invention, in a situation where the configuration parameters of the compression unit 100 and / or any other information in the information record 412 are changed, the current file 404 is closed and a new one is opened (with the same designation). As a result, the data in the header 410 and the information record 412 of the single file 404 refer to the complete file without changes. Alternatively or additionally, one or more of the fields of the record 412 comprise a portion of memory fixing various configuration values for various data records 414 and / or various signals 422.

As one option, the header 410 of the file identifies the session 402 to which the file 404 belongs, and the position of this file within the session. In some embodiments, the title also indicates the type and version by a procedure known to those skilled in the art. Alternatively or additionally, it indicates the local time at which the file data begins and / or ends. As another alternative or addition, the file header 410 captures the internal synchronization point at which the file data begins and / or ends. In some embodiments of the invention, the header 410 contains a line field that the user can use to identify the file.

Information record 412 may contain all the configuration information required to understand the materials contained in file 404. Therefore, each such file is a stand-alone unit. Alternatively, some information that usually does not change is included only in the first file 404 of each session 402. When a change occurs during the session, the changed information can be included in the file 404 in which the change occurred, or in all files 404 whose values differ from those indicated in the first file of the general session 402.

A variant is possible in which the information record 412 contains an indication of whether the file data is encoded and / or compressed, and / or indicates the type of encoding and / or compression applied. In addition, for recording 412, it is possible to include an indication of the type of product, for example, the type of equipment used, information about the network topology and / or the number of signals and / or types for which data is stored. In one embodiment, entry 412 contains various lines of common characters that provide the user with obvious general information about the data in question, such as the name of the operator and the name of the company. Record 412 may also indicate the moments of the beginning and / or end of the time period to which the 404 file relates. As one of the options, record 412 indicates events in which the user or automatically correlated the correspondence between the internal time and local time.

In some embodiments of the invention, with respect to each signal for which data is collected, information record 412 indicates its type (for example, one of the waves represented by many harmonics, a real parameter, a complex parameter, a transition parameter) and an individual signal identifier, which in the case of if necessary, the user can at least partially decrypt. For each signal, an option is provided in which the information record 412 also indicates the attributes that are applied to the signal in the method shown in FIG. 3. These attributes include parameters such as nominal value and tolerance. For each signal, alternatively or additionally, a conversion indicator and / or function are provided that are used to convert the measured value to a real amplitude value (for example, as a consequence of the conversion coefficient used).

As one of the options, the information record 412 contains, in addition, information about special events (anomalies), which is not properly provided through compressed information stored in the data records 414. For example, in some embodiments, the compression unit 100 comprises an interface suitable for identifying fast transients associated with high frequencies. Data on such anomalies can be stored in record 412 along with the time when such an anomaly occurred. In some embodiments, implementation 412 contains debugging information.

It should be noted that the data designated in this case as contained in the information record 412 can instead be partially or completely placed in the header 410 and / or in other records of the file 404. For example, the record 412 can be divided into many records related to different information.

In one embodiment, eight signals are tracked (4 for current and 4 for voltage) by measuring at a frequency of 12.6 kHz. For each measurement, for example, 2 bytes are required, i.e. every second, the selected measurements require 204.8 KB, and for six months - approximately 6400 GB. Using the compression method described above, a compression ratio of approximately 1: 1000 was achieved, and the use of additional universal compression (LZV) resulted in a total compression ratio of approximately 1: 10000.

In the above embodiments, the compression unit 100 simultaneously comprises an analog sensor 110 and a compressor 112. However, it does not have a storage unit 120, these two units are separated. In other embodiments, the analog sensor 110 and, alternatively, the ADC 114 are located in a unit separate from the compressor 112. This arrangement occurs, for example, when the measurement is carried out by means of a tracking unit equipped with a standard alarm. Such a block is modified in such a way as to ensure that the compressor 112 transmits all the measured data, and not just the data measured in the event of an emergency. Alternatively or additionally, the storage unit is placed inside the compression unit 100.

The device of the compression unit 100 is not limited to using any specific electrical wires and / or network topologies. In particular, this unit can be used in the case of any multiphase or single-phase electrical network. In addition, the measurement can be carried out for land lines, it can cover not all electric lines of the network, but only some, and / or provide any combination of current and voltage measurements. In addition to the above, the compression unit 100 can be used in the embodiment with symmetric and asymmetric circuits, and essentially with any topological network circuits, including triangle / star circuits and other hybrid circuits, as well as other connection points. Compression unit 100 can be used with substantially any voltage — low, medium, high, and ultrahigh.

Although in the above description, compression is applied to the parameters or harmonics of a single cycle of the main frequency of the network, the same compression method can be applied to parameters determined less frequently or more often. In some embodiments of the invention, the conversion is performed each time together on multiple cycles, and harmonics are compared for each multiple cycles. Alternatively, the conversion is carried out on measurements representing part of the cycle of the electric line, for example half or one sixth of the cycle. Such embodiments are used, in particular, in electrical networks, of which it is known that they are subjected to repeated exposure to a frequency different from the line frequency. In some embodiments, the conversion frequency is predetermined. Alternatively, it is controlled in accordance with the measured signals.

Alternatively or additionally, the values of one or more parameters, such as recession, increase and / or runout, are determined once every few cycles and compression is performed in accordance with the results. In addition, as an alternative or in addition, data on one or more waveforms can be stored twice, namely with high and low resolution. As one option, a high resolution copy is stored in each cycle with a relatively high compression tolerance. A low resolution copy is stored only once for every few cycles, with a much lower compression tolerance.

Alternative to storing harmonics and parameters, the compression unit 100 may store waveform configurations. Among other options, every time this parameter changes significantly (for example, by more than 5-10%), a new configuration is saved.

It should be appreciated that the methods described above can be varied in a wide variety of ways. It must also be emphasized that the above description of the methods and device must be interpreted, on the one hand, as a representation of the device for implementing the methods, and on the other hand, as ways of using the specified device.

The present invention has been presented by way of examples of a detailed description of its embodiments. These options are provided only as examples and should not be construed as limiting the scope of the invention. In particular, processor 118 need not be located in compression unit 100. The functions of the processor 118 can be performed not by means of the indicated unit, but by its other elements, such as compressor 112. It should be borne in mind that the distinguishing features and / or steps described for one embodiment of the invention can be applied to other variants, moreover, not all embodiments include a complete set of features and / or steps presented in a particular drawing or described with respect to one embodiment. For those skilled in the art, permissible changes to embodiments of the invention will be apparent.

It must be emphasized once again that some of the embodiments described above are provided as examples only and may be merely modifications that are considered by the inventors to be optimal. Therefore, they may include schemes, operations or their details, which may not be necessary for the implementation of the invention. The described schemes and operations can be replaced by known equivalents that perform the same functions, even if they differ in design or principle actions. Thus, the scope of the invention is limited only by the features and limitations given in the claims. The terms “contains”, “includes”, “has”, and their derivatives mean the wording “includes, but is not limited to,” used therein.

Claims (49)

1. A method of compressing values characterizing the waveforms of a monitored electrical energy signal, including collecting data representing periods of the electric energy signal, decomposing the waveform into a plurality of components spanning a plurality of periods of the waveform, and separately compressing the values of at least some components spanning many periods. 2. The method according to claim 1, characterized in that the decomposition of the waveform of the electrical energy signal into components includes the decomposition of the waveform of the electrical energy signal into frequency components. 3. The method according to claim 1, characterized in that the compression of the values of at least some of the components includes fitting the time segments of the components to the model and registering the fitting coefficients. 4. The method according to claim 3, characterized in that the model contains a constant time function. 5. The method according to claim 4, characterized in that the registered coefficients for the constant function, depending on time, contain a single value and length. 6. The method according to claim 3, characterized in that the model contains a monotonic time function. 7. The method according to claim 1, characterized in that the collection of data representing the periods of the signal includes collecting measurements of the electrical energy signal and dividing the measurements into groups corresponding to the cycles of the electrical energy signal. 8. The method according to claim 7, characterized in that the decomposition of the waveform includes converting the measurements of each group into values of harmonic components. 9. The method according to claim 8, characterized in that the separate compression of at least some components includes storing for each harmonic pairs of average values and the number of cycles in which the value is close to the average value. 10. The method according to claim 9, characterized in that the number of cycles in which the value is close to the average value is established by determining the minimum and maximum sequence of harmonic values and determining the moment when the difference between the minimum and maximum becomes greater than the specified difference. 11. The method according to claim 10, characterized in that the predetermined difference is expressed as a percentage of the previous average harmonic value. 12. The method according to claim 10, characterized in that the predetermined difference is expressed as a percentage of the expected harmonic value. 13. The method according to claim 7, characterized in that the collection of measurements includes sampling analog signals. 14. The method according to any one of claims 7 to 13, characterized in that the division of the measurements into groups includes a repeated determination of the fundamental frequency of the signal and a corresponding determination of the cycles of the electrical energy signal. 15. The method according to 14, characterized in that the sampling of the signals includes sampling at a frequency determined depending on the fundamental frequency of the signal. 16. The method according to 14, characterized in that the repeated determination of the fundamental frequency of the signal is made on the basis of the received measurements. 17. The method according to 14, characterized in that the repeated determination of the fundamental frequency of the signal is made on the basis of the analog signal to which the obtained measurements correspond. 18. The method according to claim 8, characterized in that the conversion of the measurements of each group includes a fast Fourier transform. 19. The method according to claim 8, characterized in that it includes the application of a lossless compression method to the compressed harmonics. 20. The method according to claim 1, characterized in that it comprises storing at least some of the compressed components in a file structure representing a plurality of electrical energy signals. 21. The method according to claim 1, characterized in that it includes the preservation of the compressed components in a file structure representing a signal of electrical energy continuously for a time interval exceeding a month. 22. The method according to claim 1, characterized in that it includes the transmission of compressed components over the communication line. 23. The method according to claim 1, characterized in that the compression of at least some components includes separate compression of each component. 24. The method according to claim 1, characterized in that the compression of at least some of the components includes real-time compression. 25. The method according to claim 1, characterized in that the compression of at least some components includes compression using lossy compression. 26. The method according to claim 1, characterized in that the separate compression of at least some components includes separate compression, covering at least three signal periods. 27. The method according to claim 1, characterized in that the electrical energy signal contains a current signal. 28. The method according to claim 1, characterized in that the electrical energy signal contains a voltage signal. 29. The method according to claim 1, characterized in that the collection of data representing the periods of the signal includes collecting data representing cycles of the waveform. 30. The method according to claim 1, characterized in that the collection of data representing periods of the signal includes collecting data representing periods shorter or longer than waveform cycles. 31. The method according to claim 1, characterized in that the decomposition of the waveform into many components is a decomposition into components that coexist in time. 32. A method of compressing values of a monitored electrical energy signal, including collecting measurements of an electrical energy signal, dividing the measurements into groups corresponding to cycles of an electrical energy signal, converting the measurements of each group into harmonic component values, and storing the display of harmonic component values in a non-volatile storage medium continuously for at least a week. 33. The method according to p, characterized in that the storage of the display includes storing a compressed display of the values of the harmonic components. 34. The method according to p. 33, wherein storing the compressed display includes storing compression based on the joint compression of the values of each harmonic component spanning many cycles. 35. An apparatus for monitoring electric energy signals, comprising an electric line interface for measuring electric line signals, a non-volatile storage medium and a processor configured to record information on the waveform of the measured electric line signals continuously into the non-volatile storage medium, regardless of whether identified anomaly. 36. The device according to p, characterized in that the signals of the electric line contain at least one current signal. 37. The device according to clause 35, wherein the signals of the electric line contain at least one voltage signal. 38. The device according to clause 35, wherein the processor is configured to record the display continuously for at least a week. 39. The device according to p. 35, characterized in that it is not adapted to identify anomalies. 40. The device according to p, characterized in that the storage medium has a capacity of less than 1 GB. 41. The device according to clause 35, wherein the storage medium has a capacity of more than 1 GB. 42. The device according to clause 35, wherein the processor is configured to compress the measured signals of electrical energy, at least with a given minimum compression ratio. 43. The device according to item 42, wherein the processor is configured to control the level of compression loss to ensure that the specified minimum compression ratio is achieved. 44. The device according to clause 35, wherein the processor is configured to compress the measured signals of electrical energy in real time. 45. The device according to clause 35, wherein the processor is configured to compress the measured signals of electrical energy according to the lossy compression method. 46. The device according to clause 35, wherein the interface is configured to receive measurements of electrical energy signals with a frequency of at least 8 measurements per cycle of electrical energy signals. 47. A method of compressing values of a monitored electrical energy signal, including collecting measurements of an electrical energy signal and compressing measurements of an electrical energy signal using a lossy compression method. 48. The method according to clause 47, wherein the compression of measurements includes compression in real time. 49. The method according to clause 47, wherein the collection of measurements includes collecting with a frequency of at least 50 measurements per cycle of a signal of electrical energy.

Images