Method and system for extremely light compression of metering data of an electric energy meter, device, medium

Abstract

The application discloses a kind of extremely light compression method and system, equipment, medium of electric energy metering data, the method is constructed by residual coding, zero dense transposition and zero byte segment compression three-level extremely light compression architecture, first by residual coding generation longitudinal height redundancy and the longitudinal zero dense sequence with zero byte mainly in redundant area, again by zero dense transposition, longitudinal height zero redundant area is converted into transverse continuous zero byte segment, create favorable conditions for subsequent zero byte segment compression, finally, by the compression processing of a large number of zero byte segments, while a small amount of non-zero byte is directly retained, so as to obtain compressed sequence, greatly improve the compression efficiency and decompression speed of electric energy metering data, and, the entire compression algorithm is composed of low-cost operations such as addition, subtraction, shift, comparison, etc., the MCU of electric energy meter can run without pressure, and the memory overhead is limited to a very small auxiliary table and a small amount of buffer area, which is well suited to the low-power environment of electric energy meter.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data compression, in particular, to an extremely lightweight compression method and system for metering data of an electric energy meter, an electronic device and a computer readable storage medium. BACKGROUND

[0002] In the embedded running environment of the electric energy meter, the system hardware resources (including processor computing power, memory capacity and storage space) are generally limited, and the existing compression algorithms usually have the following deficiencies when applied to the electric energy meter:

[0003] (1) High hardware resource occupation and large algorithm implementation complexity, which will significantly squeeze the available resources of other applications of the electric energy meter and affect the overall performance and scalability once deployed;

[0004] (2) The characteristics of the metering data of the electric energy meter (such as smooth data change trend, obvious periodicity and limited numerical range) are not fully combined, so the compression efficiency and decompression speed still need to be improved.

[0005] Therefore, it is urgent to design an extremely lightweight data compression algorithm to efficiently process and store metering data, fully utilize the statistical rules and redundancy characteristics of the electric energy meter data, realize efficient data compression and fast decompression under the premise of maintaining extremely low resource occupation and computing complexity, and meet the needs of low power consumption, real-time and long-term stable operation. SUMMARY

[0006] The present application provides an extremely lightweight compression method and system for metering data of an electric energy meter, an electronic device and a computer readable storage medium, which can construct an extremely lightweight compression architecture for metering data of an electric energy meter, realize efficient and high-quality compression of metering data of an electric energy meter, and greatly improve the compression efficiency and decompression speed.

[0007] According to one aspect of the present application, an extremely lightweight compression method for metering data of an electric energy meter is provided, comprising the following contents:

[0008] Collecting an original metering data sequence of the electric energy meter;

[0009] Residual coding the original metering data sequence to generate a longitudinal zero-dense sequence;

[0010] Zero-dense transposition of the longitudinal zero-dense sequence to convert the longitudinal zero-dense sequence into a transverse zero-dense sequence;

[0011] Compressing the zero byte segment in the transverse zero-dense sequence to obtain a compressed sequence.

[0012] Further, the process of compressing the zero byte segment in the transverse zero-dense sequence includes the following contents:

[0013] Reading bytes from the transverse zero dense sequence in sequence, if the number of continuous zero bytes L is less than a preset threshold n, the original data is retained; if the number of continuous zero bytes L is greater than or equal to the preset threshold n, a difference value R = L - n is calculated, the first n zero bytes are retained, and the coding starts from the n+1th zero byte; if R is less than or equal to 127, one byte is used for coding, the highest bit of the byte is 0, and the last seven bits are used to write the difference value R; if R is greater than or equal to 128, multiple bytes are used for coding, and in the first byte of the multiple bytes, the highest bit is 1, and the last seven bits are used to write the number of bytes consumed by the subsequent difference value R, and the remaining bytes of the multiple bytes are used to write the difference value R.

[0014] Further, the process of compressing the zero byte segment in the transverse zero dense sequence includes the following contents:

[0015] A shrink threshold is set, and bytes are read from the transverse zero dense sequence in sequence; if the length of the zero byte segment is greater than or equal to the shrink threshold, shrink is triggered, at which time the zero byte segment is pruned, and the starting position and length of the pruned segment are recorded in the shrink table.

[0016] Further, if the collected is power data or current data, the process of compressing the zero byte segment in the transverse zero dense sequence further includes the following contents before the compression:

[0017] The transverse zero dense sequence is replaced with zero bridge, and a replacement table is generated to lengthen the length of the zero byte segment in the transverse zero dense sequence.

[0018] Further, the process of replacing the transverse zero dense sequence with zero bridge and generating a replacement table includes the following contents:

[0019] The transverse zero dense sequence is scanned, and the non-zero byte segment between two adjacent zero byte segments in the transverse zero dense sequence is regarded as a zero bridge; if the length of the zero bridge is less than or equal to a preset threshold, the zero bridge is replaced with a zero byte segment and the byte mode of the zero bridge is determined; if the byte mode of the zero bridge is a set mode, the starting position, byte mode and length of the zero bridge are recorded in the replacement table; if the byte mode of the zero bridge is an extended mode, the starting position, byte mode, length and original byte sequence of the zero bridge are recorded in the replacement table.

[0020] Further, the process of transposing the longitudinal zero dense sequence with zero dense includes the following contents:

[0021] First, the data unit in the longitudinal zero dense sequence is taken as the granularity, each byte is expanded by bit, from the most significant bit to the least significant bit, into 8 independent bits, and then re-serializes according to the bit depth first principle to generate a recombined continuous data stream;

[0022] Or, the longitudinal zero-dense sequence is first converted into an N*M matrix, where N represents the number of time points, and M represents the number of bytes contained in a data unit of the electrical quantity measurement data, and then the matrix is sequentially unfolded according to the column-first principle to generate a continuous data stream.

[0023] Further, in the process of residual coding of the original measurement data sequence, if the voltage data is collected, the first-order difference method is used to calculate the residual value, and if the power data or the current data is collected, any one of the first-order difference method, the two-point linear extrapolation method and the three-point median filter prediction method is used to calculate the residual value.

[0024] In addition, the application also provides a very light compression system for meter measurement data, comprising:

[0025] The data acquisition module is configured to acquire an original measurement data sequence of the electric energy meter.

[0026] The residual coding module is configured to perform residual coding on the original measurement data sequence to generate a longitudinal zero-dense sequence.

[0027] The zero-dense transposition module is configured to perform zero-dense transposition on the longitudinal zero-dense sequence to convert the longitudinal zero-dense sequence into a transverse zero-dense sequence.

[0028] The compression processing module is configured to perform compression processing on the zero-byte segment in the transverse zero-dense sequence to obtain a compressed sequence.

[0029] In addition, the application also provides an electronic device comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to execute the steps of the method described above by calling the computer program stored in the memory.

[0030] In addition, the application also provides a computer-readable storage medium for storing a computer program for very light compression of meter measurement data, wherein the computer program executes the steps of the method described above when running on a computer.

[0031] The application has the following advantages:

[0032] The extremely lightweight compression method of the electric energy metering data of the application first utilizes the stationarity and trend of the metering data to transform the original metering data into a large number of 0 or decimal value residuals, so that the value range of the data sequence is greatly reduced, the repetition degree is greatly increased, and the statistical entropy is significantly reduced, creating an ideal input of "low entropy and high redundancy" for coding, greatly simplifying the subsequent coding, improving the compression ratio as a whole, and converting each residual into at least one code word in the non-negative integer domain through coding, which not only preserves the positive and negative information, but also avoids supplementary code storage, greatly reducing the calculation complexity during decoding, and very suitable for the computing power and RAM constraints of the electric energy meter MCU. Then, in the longitudinal zero-dense sequence generated by residual coding, the longitudinal redundancy is significantly higher than the transverse redundancy, and the high redundancy area in the longitudinal direction is mainly composed of zero bytes, so after zero-dense transposition of the longitudinal zero-dense sequence, the longitudinal zero redundancy is converted into a transverse continuous zero byte segment, creating conditions for subsequent zero byte segment compression processing. Finally, a large number of zero byte segments in the transverse zero-dense sequence are compressed, and a small amount of non-zero bytes are directly retained, so as to obtain a compressed sequence, realizing efficient and high-quality compression of electric energy metering data, greatly improving the compression efficiency and decompression speed. The entire compression algorithm is composed of low-cost operations such as addition, subtraction, shift, comparison, etc., and for the first time, an extremely lightweight compression architecture for electric energy metering data is constructed, which can be run without pressure by the MCU of the electric energy meter, and the memory overhead is limited to a small amount of auxiliary table and a small amount of buffer area, which can be well applied to the low computing power environment of the electric energy meter.

[0033] In addition, the extremely lightweight compression system of the electric energy metering data of the application also has the above advantages.

[0034] In addition to the purposes, features and advantages described above, the application has other purposes, features and advantages. The application will be further described below with reference to the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0035] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application, illustrate the preferred embodiments of the application, and explain the application without imposing undue limitation on the application. In the drawings:

[0036] Figure 1 is a flowchart of the extremely lightweight compression method of the electric energy metering data of the preferred embodiment of the application;

[0037] Figure 2 is a schematic diagram of a certain area measured power data sequence in the preferred embodiment of the application;

[0038] Figure 3 is a schematic diagram of a certain area measured voltage data sequence in the preferred embodiment of the application;

[0039] Figure 4is a schematic diagram of a certain area measured current data sequence in the preferred embodiment of the present application;

[0040] Figure 5 is a schematic diagram of a certain area measured current data sequence in the preferred embodiment of the present application; Figure 2 is a schematic diagram of a longitudinal zero dense sequence generated after first-order difference processing and Zigzag encoding of the power data sequence in the preferred embodiment of the present application;

[0041] Figure 6 is a schematic diagram of a certain area measured current data sequence in the preferred embodiment of the present application; Figure 2 is a schematic diagram of a residual coding sequence generated after first-order difference processing of the power data sequence in the preferred embodiment of the present application using a sign-amplitude coding method;

[0042] Figure 7 is another flowchart of the extremely lightweight compression method for electric energy metering data in the preferred embodiment of the present application;

[0043] Figure 8 is a schematic diagram of the module structure of the extremely lightweight compression system for electric energy metering data in another embodiment of the present application. DETAILED DESCRIPTION

[0044] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0045] With reference to Figure 1 , the preferred embodiment of the present application provides an extremely lightweight compression method for electric energy metering data, including the following contents:

[0046] Step S1: collecting a raw metering data sequence of an electric energy meter;

[0047] Step S2: performing residual coding on the raw metering data sequence to generate a longitudinal zero dense sequence;

[0048] Step S3: performing zero dense transposition on the longitudinal zero dense sequence to convert the longitudinal zero dense sequence into a transverse zero dense sequence;

[0049] Step S4: performing compression processing on the zero byte segment in the transverse zero dense sequence to obtain a compressed sequence.

[0050] It can be understood that the extremely lightweight compression method of the electric energy metering data of the embodiment first utilizes the stationarity and trend of the metering data, transforms the original metering data into a large number of 0 or small residual values, greatly reduces the value range of the data sequence, greatly increases the repetition degree, and significantly reduces the statistical entropy, creating an ideal input of "low entropy and high redundancy" for coding, greatly simplifying the subsequent coding, improving the compression ratio as a whole, and converting each residual into at least one code word in the non-negative integer domain through coding, which not only preserves the positive and negative information, but also avoids supplementary code storage, greatly reducing the calculation complexity during decoding, and very suitable for the computing power and RAM constraints of the electric energy meter MCU. Then, in the longitudinal zero-dense sequence generated by residual coding, the longitudinal redundancy is significantly higher than the transverse redundancy, and the highly redundant area in the longitudinal direction is mainly composed of zero bytes, and after zero-dense transposition of the longitudinal zero-dense sequence, the longitudinal zero redundancy is converted into a transverse continuous zero byte segment, creating conditions for subsequent zero byte segment compression processing. Finally, by compressing a large number of zero byte segments in the transverse zero-dense sequence, and directly retaining a small number of non-zero bytes, a compressed sequence is obtained, realizing efficient and high-quality compression of electric energy metering data, greatly improving compression efficiency and decompression speed. Moreover, the entire compression algorithm is composed of low-cost operations such as addition, subtraction, shifting, and comparison, and for the first time, an extremely lightweight compression architecture for electric energy metering data is constructed, which can be run by the MCU of the electric energy meter without any pressure, and the memory overhead is limited to a small amount of auxiliary table and a small amount of buffer area, which can be well applied to the low-computing-power environment of the electric energy meter.

[0051] In the step S1, the type of the original metering data of the electric energy meter specifically includes power data, voltage data and current data, and the application preferably compresses the power data and the current data. In order to improve the storage efficiency, operation efficiency and compression efficiency, the original metering data is usually recorded in the form of fixed-point scaling and integer storage, specifically, the original decimal is enlarged by a fixed scaling factor to convert it into an integer, and then subsequent storage, calculation and compression processing are completed, which can maintain the required precision while avoiding resource overhead and platform differences caused by floating-point format. Specifically, for power data, a single power value is regarded as a data unit, 1 decimal place (unit W) is retained, and after being enlarged by 10 times, it is stored as a 4-byte integer; for voltage data, a single voltage value is regarded as a data unit, 1 decimal place (unit V) is retained, and after being enlarged by 10 times, it is stored as a 2-byte integer; for current data, a single current value is regarded as a data unit, 3 decimal places (unit A) are retained, and after being enlarged by 1000 times, it is stored as a 4-byte integer; therefore, the original metering data sequence is composed of a certain number of data units arranged in time sequence. For example, as shown in FIG. 1, the original metering data sequence is composed of a plurality of data units arranged in time sequence, and each data unit is composed of a plurality of data elements, and each data element is composed of a plurality of bytes. Figure 2The certain area measured power data sequence shown in the figure containing 16 data units adopts little-end byte sequence, the first active power data unit, i.e. field 0x00005250, corresponds to the actual value 2107.2W, and the calculation process is: 0x5250 / 10=2107.2W, and the power data sequence is composed of the above four byte data units; as shown in the figure Figure 3 The certain area measured voltage data sequence shown in the figure containing 16 data units adopts little-end byte sequence, the first voltage data unit, i.e. field 0x08BF, corresponds to the actual value 223.9V, and the calculation process is: 0x08BF / 10=223.9V; as shown in the figure Figure 4 The certain area measured current data sequence shown in the figure containing 16 data units adopts little-end byte sequence, the first current data unit, i.e. field 0x00003833, corresponds to the actual value 14.387A, and the calculation process is: 0x00003833 / 1000=14.387A.

[0052] It can be understood that, from the area measured data Figures 2-4 It can be understood that, from the area measured data

[0053] Among them, the residual calculation of the original metering data sequence is to generate a predicted value according to the predictor, and then subtract the actual value from the predicted value to obtain the residual, and the residual calculation method can adopt any one of the first-order difference method, the two-point linear extrapolation method and the three-point median filtering prediction method, wherein the first-order difference method is to predict the value of the next point with the current value, the two-point linear extrapolation method is to linearly predict the value of the next data point with the latest two data points, and the three-point median filtering prediction method is to predict the value of the next data point with the value of the middle one among the latest three data points. Specifically, the coding rule of the first-order difference method is: the seed is x1(t=1), the predictor is , and the residual is The encoding rule of the two-point linear extrapolation method is that the seed stores x1, x2 (t = 1, 2), the predictor is , and the residual error is The encoding rule of the three-point median filter prediction method is that the seed stores x1, x2, x3 (t = 1, 2, 3), the predictor is , and the residual error is .

[0054] For example, assuming that the original measurement data sequence is x = [1000, 1000, 1000, 1001, 1200, 1210, 1220, 1230, 1240, 1400, 1240, 1250], the results of calculating the residual error values by using the above three residual error calculation methods are shown in Table 1.

[0055] Table 1. Calculation results of different residual error calculation methods

[0056]

[0057] As can be seen from Table 1, the first-order difference method has the simplest and lightest operation, almost no additional operation burden, is extremely friendly to the platform segment and small fluctuations (such as voltage data), and the residual error is mostly 0 or very small, for example, the residual error of the platform segment (t = 2-3) is 0, when a step change is encountered, only a large residual error occurs at the transition point, for example, the residual error of the step point (t = 5) and the spike (t = 10, 11) is large, and the remaining points remain stable. The two-point linear extrapolation method is suitable for data sequences with linear trends (such as power / current slowly rising or falling in the steady state stage), and the residual error is almost 0 in the trend segment, for example, the residual error of the linear climbing segment (t = 7-9) is completely zero, which can eliminate linear changes more than the first-order difference. However, in the case of sudden steps or spikes, a large residual error will occur near the transition point, for example, the step (t = 5, 6) and the spike (t = 11). The three-point median filter prediction method has the strongest ability to resist spikes and burrs (such as occasional abnormal points in current or power), for example, the residual error of the spike (t = 11) is 0, and for local outliers, the median filter prediction can accurately fit the true trend, making the residual error very small or even 0, and the trend segment may have a slight lag, and the residual error is slightly larger than that of linear extrapolation, for example, the residual error of the linear climbing segment (t = 7-9) remains at +20, but it has stronger robustness.

[0058] Therefore, in the process of residual error encoding of the original measurement data sequence, if voltage data is collected, the first-order difference method is preferably used to calculate the residual error value, and if power data or current data is collected, any one of the first-order difference method, the two-point linear extrapolation method, and the three-point median filter prediction method is used to calculate the residual error value.

[0059] In addition, after the residual calculation on the original metering data sequence, a residual data sequence can be obtained, and the residual data sequence is encoded to represent small values and zeros more efficiently. For example, the residual data sequence can be encoded by Zigzag encoding, which can be expressed as: The Zigzag encoding uses low bit to store the sign and high bit to store the amplitude, which can map the small signed integer to the same small non-negative integer domain, thereby generating the residual sequence, i.e., the vertically zero-dense sequence, for example, Figure 2 After the first-order difference processing and Zigzag encoding of the power data sequence in Figure 5 , the generated vertically zero-dense sequence is shown in Figure 5 . It can be seen that there are continuous 0x00 in the vertically zero-dense sequence.

[0060] Alternatively, the residual data sequence is encoded by the sign-amplitude encoding method. Specifically, the encoding process uses the most significant bit (MSB) to store the sign and the remaining bits to store the residual absolute value, i.e., the MSB as the sign bit, storing the sign, 0 indicating non-negative and 1 indicating negative value, and the remaining bits storing the residual absolute value. For example, for the power data sequence in Figure 2 , the first frame power data P0=0x00005250, the second frame power data P1=0x00005269, and the first residual data r1=P1-P0=0x00000019 if the first-order difference is used to calculate the residual value. The MSB sign bit is 0 and the amplitude height is 0x19, i.e., corresponding to 19 00 00 00 in Figure 6 . In addition, the third frame power data P2=0x00005263, and the second residual data r2=P2-P1=-0x00000006. The MSB sign bit is 1 and the amplitude height is 0x00000006, i.e., corresponding to 06 00 00 80 in Figure 6 . In addition, in Figure 6 , when the last byte of each data unit is 00, it indicates that the sign bit is non-negative, and when it is 80, it indicates that the sign bit is negative. In addition to the MSB sign bit and the residual absolute value, there are a large number of repeated bits (0) in the vertical bit column.

[0061] It is understood that this invention uses an encoding method that stores the symbol in the most significant bit and the absolute value of the residual in the remaining bits. This method converts each residual data into at least one codeword in the non-negative integer field. For example, when the difference between two adjacent frames of data is small, such as less than or equal to 255, each residual data is converted into one codeword. When the difference between two adjacent frames of data is large, such as greater than 255, each residual data is converted into two codewords. This preserves both positive and negative information and avoids the need for two's complement storage. The high bits remain 0xFF or 0x00 for a long time, and the equivalent entropy does not actually decrease. At the same time, this symbol-amplitude format only requires one bit operation to recover the original symbol during decoding, resulting in extremely low computational complexity. This is very suitable for the computing power and RAM constraints of the energy meter MCU.

[0062] Furthermore, in step S3, since the vertically zero-dense sequence contains continuously distributed zero-byte segments in the vertical direction, transposing these continuously distributed 0x00 segments horizontally creates conditions for subsequent zero-byte segment compression processing. The process of transposing the vertically zero-dense sequence includes the following:

[0063] First, taking the data units in the vertical zero-dense sequence as the granularity, each byte is expanded bit by bit, and then split into 8 independent bits from the most significant bit to the least significant bit. Then, they are re-concatenated according to the bit depth priority principle to generate a recombined continuous data stream, which is the horizontal zero-dense sequence.

[0064] Specifically, taking the data units in the vertical zero-dense sequence as the granularity, each byte bit is expanded bit by bit, from the most significant bit (MSB) to the least significant bit (LSB) into 8 independent bits, which can be represented as: ,in, b t,i Indicates the first t The first time point i The value of each byte. Then, the vertical zero-dense sequence is no longer arranged in chronological order, but is re-concatenated according to the bit depth priority principle, that is, the 7th bit of all data units is written first, then the 6th bit, ..., until the 0th bit, which can be represented as: N represents the number of time points, and M represents the number of bytes contained in the data unit. For example, taking a 4-byte power data unit as an example, the reassembled data stream is: {7th bit of column 0, 7th bit of column 1, ..., 7th bit of column 3; 6th bit of column 0, 6th bit of column 1, ..., 6th bit of column 3; ...; 0th bit of column 0, 0th bit of column 1, ..., 0th bit of column 3}.

[0065] It is understandable that, since the vertical zero-dense sequence exhibits extremely strong stability in the vertical bit dimension: the high bits are almost always 0 and the low bits often repeat between adjacent samples, this application uses bit column recombination and vertical to horizontal mapping to squeeze the highly stable bit column to adjacent positions, and instantaneously constructs a large number of locally repeating 0x00 in the horizontal direction, which creates favorable conditions for subsequent zero-byte segment compression processing.

[0066] Optionally, the process of performing a zero-dense transpose on the vertical zero-dense sequence may also include the following:

[0067] First, the vertical zero-dense sequence is converted into an N×M matrix, where N represents the number of time points and M represents the number of bytes contained in the data unit of the original measurement data. Then, the matrix is ​​expanded in order according to the column priority principle to generate a continuous data stream, which yields the horizontal zero-dense sequence.

[0068] Specifically, a vertical zero-dense sequence containing records at N time points can be converted into an N×M matrix, where N represents the number of time points and M represents the number of bytes contained in a data unit of the original measurement data. For example, for power data, M is 4, and the matrix can be represented as follows: , b t,i Indicates the first t The first time point i A value of byte, 0≤ t ≤ N -1, 0≤ i ≤ M-1 Expanding the matrix into a continuous data stream in column-major order, it can be represented as: Thus, a horizontally zero-dense sequence is obtained.

[0069] It is understandable that the byte-level reassembly in this application is based on bytes, and it can also achieve vertical to horizontal mapping. Similarly, it can instantaneously construct a large number of locally repeating 0x00s in the horizontal direction, creating favorable conditions for subsequent zero-byte segment compression processing. However, compared with the bit-level reassembly mentioned above, byte-level reassembly constructs relatively fewer zero-byte segments, and the compression efficiency is relatively lower, but its implementation is simpler. Bit-level reassembly, on the other hand, introduces more bit operations and boundary processing, making its implementation relatively more complex.

[0070] It is understood that after the zero-dense transpose, the vertical zero-dense sequence has been converted into a horizontal zero-dense sequence. Therefore, in step S4, the compressed sequence can be obtained by compressing the zero-byte segments (i.e., continuously distributed 0x00) in the horizontal zero-dense sequence and retaining the other non-zero sub-stages.

[0071] The compression process of the zero byte segment in the transverse zero dense sequence includes the following contents:

[0072] The bytes are sequentially read from the transverse zero dense sequence. If the number of continuous zero bytes L is less than the preset threshold n, the original data is reserved. If the number of continuous zero bytes L is greater than or equal to the preset threshold n, the difference value R=L-n is calculated, the first n zero bytes are reserved, and the (n+1)th zero byte is encoded. If R is less than or equal to 127, one byte is used for encoding, the highest bit of the byte is 0, and the last seven bits are used to write the difference value R. If R is greater than or equal to 128, multiple bytes are used for encoding, and in the first byte of the multiple bytes, the highest bit is 1, and the last seven bits are used to write the number of bytes consumed by the subsequent difference value R. The remaining bytes of the multiple bytes are used to write the difference value R.

[0073] Specifically, the bytes are sequentially read from the transverse zero dense sequence. If it is not a zero byte (i.e. 0x00), it is directly output as is. If it is a zero byte, the length L of the continuous zero byte is counted. If the length L of the continuous zero byte is less than the preset threshold n, the original data is reserved, i.e. the L zero bytes are output as is. If the length L of the continuous zero byte is greater than or equal to the preset threshold n, the difference value R=L-n is calculated, the first n zero bytes are reserved, and the (n+1)th zero byte is encoded. If R≤127, one byte is used for encoding, the highest bit of the byte is 0, which belongs to a flag bit, and 0 indicates the number flag, indicating that the last seven bits of the byte correspond to the number of remaining zero bytes, i.e. the value of R. When the byte is 0x00, it indicates that the value of the last seven bits is 0, i.e. R=0. If R≥128, multiple bytes are used for encoding, wherein the highest bit of the first byte is 1, indicating the occupied byte flag, i.e. the last seven bits of the first byte indicate the number of bytes consumed by R, and the remaining bytes are used to write the difference value R. In addition, if no other instructions are given, the bytes of the present application are in little-endian byte order. For example, assuming n=2, 0x00 0x01 is encoded as 0x000x01; 0x00 0x00 0x01 is encoded as 0x00 0x00 0x00 0x01; 0x00 0x00 0x00 0x01 is encoded as 0x000x00 0x01 0x01; 0x00 0x00 0x00 (a total of 5 0x00) 0x01 is encoded as 0x00 0x00 0x03 0x01; 0x000x00 0x00 (a total of 502 0x00) 0x01 is encoded as 0x00 0x00 0x82 0xf4 0x01 0x01.

[0074] It can be understood that the encoding process of the compression process described above can be represented as:

[0075] 1) reading bytes from the input data stream in sequence;

[0076] 2) If not 0x00, output as is;

[0077] 3) If 0x00, count the length of consecutive zero bytes L;

[0078] 4) If L < n, output these L zero bytes as is;

[0079] 5) If L ≥ n:

[0080] a) output n 0x00;

[0081] b) read 1 length descriptor byte: ;

[0082] c) output R 0x00;

[0083] i. R ≤ 127: write 1 byte (MSB=0, low 7 bits written value R);

[0084] ii. R ≥ 128: write 1 header byte (MSB=1, low 7 bits k, indicating the number of bytes of R following), and then write k bytes of R in little endian;

[0085] 6) Continue processing the subsequent data until the end.

[0086] In addition, the decoding process can be represented as:

[0087] 1) Read bytes from the encoded data stream in order;

[0088] 2) If not 0x00, output as is;

[0089] 3) If 0x00, read the number of consecutive zero bytes c (at most n);

[0090] 4) If c < n, output these c zero bytes;

[0091] 5) If c = n:

[0092] a) output n 0x00;

[0093] b) read 1 length descriptor byte:

[0094] i. < 0x80: R = value;

[0095] ii. ≥ 0x80: low 7 bits = k, and then read k bytes (little endian) to get R;

[0096] c) output R 0x00;

[0097] 6) Continue processing the subsequent data until the end.

[0098] It can be understood that the application can efficiently compress the highly redundant and repetitive zero bytes by the above-mentioned encoding compression of the zero byte segment, and the algorithm is entirely composed of low-cost operations such as addition, subtraction, shift and comparison, and has the advantages of ultra-low computational complexity, extremely small memory occupation, and almost instantaneous processing speed.

[0099] As preferred, the process of compressing the zero byte segment in the transverse zero dense sequence can also include the following contents:

[0100] A shrinkage threshold is set, and bytes are sequentially read from the transverse zero dense sequence, and if the length of the zero byte segment is greater than or equal to the shrinkage threshold, shrinkage is triggered, at which time the zero byte segment is pruned, and the start position and length of the pruned segment are recorded in the shrinkage table.

[0101] Specifically, a shrinkage threshold T is set, and bytes are sequentially read from the transverse zero dense sequence, and when encountering zero bytes that are continuously distributed to form a zero byte segment, the length L of the zero byte segment is counted, and if L≥T, shrinkage is triggered, at which time the zero byte segment is pruned as a whole, and a compressed sequence is obtained, and the start position and length of the pruned segment are recorded in the shrinkage table. When decompressing, the same number of 0x00 is filled back at the corresponding position according to the shrinkage table, and the original sequence can be restored.

[0102] It can be understood that the application only needs to record the start position and length of the zero byte segment by directly performing a simple hole filling / restoration on the zero byte segment of the transverse zero dense sequence, only a small shrinkage table is saved, and there is no need for a complex dictionary or entropy table, and the application has the advantages of ultra-low computational complexity, extremely small memory occupation, and almost instantaneous processing speed.

[0103] It can be understood that the general compression algorithm in the field is the Deflate algorithm, which combines two types of technologies: LZ77 dictionary compression and Huffman entropy encoding. Among them, the LZ77 dictionary compression is to find repeated segments in the data stream through a sliding window and replace the repeated content in the form of "length + offset", and the Huffman entropy encoding is to assign shorter code words to symbols with high frequency of occurrence to further improve the compression ratio. This combination ensures that the Deflate algorithm has good universality and compression ratio on various data, and is therefore widely used in formats such as ZIP and PNG. However, in the embedded scenario of electric energy meters, the Deflate algorithm has the following defects:

[0104] 1) High computational complexity: the search of the sliding window and the construction of the Huffman table require a large amount of calculation, and the compression and decompression speed is limited;

[0105] 2) Large memory consumption: Even if a small window is set, the dictionary and statistical information still need to be maintained, which is still a great pressure on low resource environments such as MCUs.

[0106] 3) Poor real-time performance: The compression rate depends on a large range of context information, which leads to the algorithm having to cache, search and encode, making it difficult to process data in real time.

[0107] Therefore, the Deflate algorithm is more suitable for embedded applications with medium or above computing power and memory resources, such as high-end Internet of Things (IoT) electricity meters, which are usually equipped with about 1 MB of RAM and a 192 MHz MCU clock speed. In such operating environments, the computing and memory overhead of the Deflate algorithm is still acceptable. However, the most commonly used ordinary electricity meters (such as typical 20-type electricity meters) often only have less than 100 KB of RAM and a MCU clock speed of about 24 MHz. In such a restricted environment, the high resource requirements of the Deflate algorithm are obviously not suitable.

[0108] Therefore, the two zero-byte segment compression algorithms proposed in this application do not require sliding window searching and entropy table construction, only a small auxiliary table needs to be saved, and a large dictionary is not required. These algorithms have the advantages of ultra-low computational complexity, extremely small memory occupation, and near-instantaneous processing speed. Data can be compressed as soon as it arrives, making them suitable for streaming and embedded scenarios. Moreover, given the smooth, step-like and sparse noise characteristics of electricity metering data, these algorithms greatly improve the compression rate and compression efficiency, and can be well applied to ordinary electricity meters with limited computing power.

[0109] Optionally, as shown in Figure 7 If power data or current data is collected, the extremely lightweight compression method for electricity metering data further includes the following steps before compressing the zero-byte segments in the transverse zero-dense sequence:

[0110] Step S34: Perform zero bridge replacement on the transverse zero-dense sequence and generate a replacement table to lengthen the zero-byte segment length in the transverse zero-dense sequence.

[0111] It can be understood that, for power data and current data, most of the time presents a smooth change, only in the state switching of the appliance start-stop, etc. The jump value appears, and such jump value often presents a longitudinal extremely sparse high byte outlier after residual coding, and the voltage data is very stable and basically will not appear such sparse outlier byte. After zero dense transposition, it corresponds to the bridge of the "broken" zero byte segment in the horizontal direction. Therefore, the core idea of the zero bridge replacement in the application is: regarding these sparse outliers as the bridge (also called zero bridge) between zero byte segments, and replacing it with zero byte 0x00 in situ, so as to splice into longer 0x00 run, thereby the length of the zero byte segment in the horizontal zero dense sequence can be lengthened, which is beneficial to further improve the compression efficiency of the subsequent zero byte segment, and a replacement table is generated to record the position and original byte of the replacement, so as to restore the original data losslessly in decoding.

[0112] The process of replacing the zero bridge of the horizontal zero dense sequence and generating the replacement table comprises the following contents:

[0113] The horizontal zero dense sequence is scanned, and the non-zero byte segment between two adjacent zero byte segments in the horizontal zero dense sequence is regarded as a zero bridge. If the length of the zero bridge is less than or equal to a preset threshold, the zero bridge is replaced by a zero byte segment and the byte mode of the zero bridge is judged. If the byte mode of the zero bridge is a set mode, the starting position, byte mode and length of the zero bridge are recorded in the replacement table. If the byte mode of the zero bridge is an extended mode, the starting position, byte mode, length and original byte sequence of the zero bridge are recorded in the replacement table.

[0114] Specifically, the horizontal zero dense sequence is scanned, and the non-zero byte segment between two adjacent zero byte segments in the horizontal zero dense sequence is regarded as a zero bridge. If the length L of the zero bridge is less than or equal to a preset threshold L max, means that the zero bridge is a small amount of outliers, the zero bridge is replaced by zero byte segment, and the byte mode of the zero bridge is determined, if the byte mode of the zero bridge is a set mode, that is, the zero bridge is composed of a single byte, and the byte belongs to a predefined set, such as any one of 0x01, 0x02 and 0x03, etc., the starting position, byte mode and length of the zero bridge are recorded in the replacement table, if the zero bridge cannot be expressed by the set mode, the extended mode is used to express, that is, the zero bridge is not composed of a single byte, or the byte in the zero bridge does not belong to the predefined set, the starting position, byte mode, length and original byte sequence of the zero bridge are recorded in the replacement table. In order to realize compact storage, the byte mode and length are uniformly encoded by a byte, wherein the high 2 bits are used to represent the byte mode, for example, 00, 01, 11 represent 0x01, 0x02, 0x03 in the set mode respectively, and 10 represents the extended mode; the low 6 bits are used to write the length of the zero bridge. For example, for the transverse zero dense sequence: 0x00 0x00 0x01 0x00 0x00, the 3rd byte is scanned to be: 0x01, which meets the replacement condition, and is replaced by 0x00, and the replacement table records (3, mode=00, length=1 (the byte code is 00000001B, that is, 0x01)). The replaced data stream is 0x00 0x00 0x00 0x00 0x00; when decompressing, (3, mode=00, length=1) is read out from the replacement table, then 1 0x01 is written back to position 3, and the original data stream can be recovered. For example, for the transverse zero dense sequence: 0x00 0x00 0x01 0x02 0x00 0x00, the 3rd~4th byte (i.e. the zero bridge) is 0x010x02, the length is 2, which is replaced by 0x00 0x00, and the replacement table records (3; mode=10, length=2 (the byte code is 10000010B, that is, 0x82); original byte=[0x01, 0x02]), the replaced data stream is 0x00 0x00 0x00 0x000x00 0x00; when decompressing, (3, mode=10, length=2, [0x01, 0x02]) is read from the replacement table, then 0x01, 0x02 are written back to positions 3, 4 in turn, and the original data stream can be recovered.

[0115] It can be understood that the application considers that the power data and the current data tend to exhibit the characteristics of longitudinal zero density and transverse sparse distribution after residual coding, but a small amount of jumps will generate high byte isolated points that "break" the zero section after zero density transposition. However, the traditional compression algorithm often processes these isolated points together with normal data, and cannot fully exert the compression potential of the long zero byte section. Therefore, the application repositions these isolated points as zero bridges, integrates them into the zero byte section through a replacement mechanism, and assists with a replacement table to ensure lossless recovery. The zero bridge replacement in the application is a special processing mechanism for sparse outlier bytes in the transverse zero dense sequence, which can further improve the compression efficiency at very low calculation and storage cost.

[0116] In addition, as Figure 8 indicated, another embodiment of the application also provides a very lightweight compression system for electric energy metering data, preferably using the very lightweight compression method for electric energy metering data as described above, comprising:

[0117] a data acquisition module for acquiring an original metering data sequence of an electric energy meter;

[0118] a residual coding module for performing residual coding on the original metering data sequence to generate a longitudinal zero dense sequence;

[0119] a zero dense transposition module for performing zero dense transposition on the longitudinal zero dense sequence to convert the longitudinal zero dense sequence into a transverse zero dense sequence;

[0120] a compression processing module for performing compression processing on zero byte sections in the transverse zero dense sequence to obtain a compressed sequence.

[0121] It can be understood that the extremely lightweight compression system of the electric energy metering data in the embodiment first utilizes the stationarity and trend of the metering data to transform the original metering data into a large number of 0 or small residual values, so that the value range of the data sequence is greatly reduced, the repetition degree is greatly increased, and the statistical entropy is significantly reduced, creating an ideal input of "low entropy and high redundancy" for coding, greatly simplifying the subsequent coding, improving the compression ratio as a whole, and converting each residual into at least one code word in the non-negative integer domain through coding, which not only preserves the positive and negative information, but also avoids supplementary code storage, greatly reducing the calculation complexity during decoding, and very suitable for the computing power and RAM constraints of the electric energy meter MCU. Then, in the longitudinal zero-dense sequence generated by residual coding, the longitudinal redundancy is significantly higher than the transverse redundancy, and the highly redundant area in the longitudinal direction is mainly composed of zero bytes, so that the longitudinal zero redundancy is converted into a transversely continuous zero byte segment after zero-dense transposition, creating conditions for subsequent zero byte segment compression processing. Finally, a large number of zero byte segments in the transverse zero-dense sequence are compressed, and a small amount of non-zero bytes are directly retained, so as to obtain a compressed sequence, realizing efficient and high-quality compression of electric energy metering data, greatly improving the compression efficiency and decompression speed. Moreover, the entire compression algorithm is composed of low-cost operations such as addition, subtraction, shift, comparison, etc., which first constructs an extremely lightweight compression architecture for electric energy metering data, and the MCU of the electric energy meter can run without pressure, the memory overhead is limited to a small amount of auxiliary table and a small amount of buffer area, and can be well applied to the low computing power environment of the electric energy meter.

[0122] In addition, the extremely lightweight compression system of the electric energy metering data further comprises:

[0123] The zero bridge replacement module is configured to perform zero bridge replacement on the transverse zero-dense sequence and generate a replacement table, so as to lengthen the length of the zero byte segment in the transverse zero-dense sequence.

[0124] It can be understood that each module of the system embodiment corresponds to each step of the method embodiment described above, so the specific working process and principle of each module will not be described here, and the corresponding reference to each step of the method embodiment described above can be made.

[0125] In addition, another embodiment of the present application also provides an electronic device comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to execute the steps of the method described above by calling the computer program stored in the memory.

[0126] In addition, another embodiment of the present application also provides a computer readable storage medium for storing a computer program for extremely lightweight compression of electric energy metering data, wherein the computer program executes the steps of the method described above when running on a computer.

[0127] The computer-readable media of general storage, in some embodiments, are any media to which the computer can read. Such computer-readable media include a floppy disk, a flexible disk, a hard disk, a magnetic tape, any other magnetic medium, a CD-ROM, any other optical medium, punch cards, paper tape, any other physical medium with patterns of holes, a RAM, a programmable ROM (PROM), an erasable programmable ROM (EPROM), a FLASH-EPROM, any other memory chip or cartridge, or any other medium from which a computer can read. The instructions can further be transmitted or received using a transmission medium. The term "transmission medium" shall include any intangible or tangible medium that is made use of to store, encode or carry the instructions for execution by a machine and includes digital or analog communications signals or other intangible media to facilitate communication of such instructions. Transmission media include coaxial cables, copper wire and fiber optics, including the wires that comprise bus that carry digital data signals.

[0128] Those skilled in the art will appreciate that embodiments of the present application can be devised for a method, a system, or a computer program product. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer readable program code thereon for use by or in connection with an instruction execution system. Program code embodied on one or more computer-usable storage media can be downloaded over a network from one computer to another computer and streamed for execution. The program code can be embodied in any form of computer program product, such as a computer program propagated in a computer

[0129] The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified by the flowchart block or blocks. Figure 1 The flowchart and / or block diagram in the drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart and / or block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions ("instruction"). It should also be noted that each block of the flowchart and / or block diagrams and Figure 1 The flowchart and / or block diagram in the drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart and / or block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions ("instruction"). It should also be noted that each block of the flowchart and / or block diagrams and

[0130] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0131] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0132] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0133] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

[0134] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for extremely lightweight compression of electricity meter metering data, characterized in that, Includes the following: Collect the raw metering data sequence from the electricity meter; Residual coding is performed on the original measurement data sequence to generate a vertical zero-dense sequence; The vertical zero-dense sequence is transposed to convert it into a horizontal zero-dense sequence. The zero-byte segments in the horizontal zero-dense sequence are compressed to obtain the compressed sequence; If the collected data is power data or current data, the following content is included before compressing the zero-byte segments in the lateral zero-dense sequence: Zero-bridge replacement is performed on the horizontal zero-dense sequence, and a replacement table is generated to extend the length of the zero-byte segment in the horizontal zero-dense sequence; The process of performing zero-bridge replacement on the lateral zero-dense sequence and generating a replacement table includes the following: The horizontal zero-dense sequence is scanned. The non-zero byte segment between two adjacent zero byte segments in the horizontal zero-dense sequence is regarded as a zero bridge. If the length of the zero bridge is less than or equal to a preset threshold, the zero bridge is replaced with a zero byte segment and the byte mode of the zero bridge is determined. If the byte mode of the zero bridge is the set mode, the starting position, byte mode and length of the zero bridge are recorded in the replacement table. If the byte mode of the zero bridge is the extended mode, the starting position, byte mode, length and original byte sequence of the zero bridge are recorded in the replacement table.

2. The method for ultra-lightweight compression of electricity meter metering data as described in claim 1, characterized in that, The process of compressing zero-byte segments in a horizontally zero-dense sequence includes the following: Read bytes sequentially from the horizontal zero-dense sequence. If the number of consecutive zero bytes L is less than a preset threshold n, then retain the original data. If the number of consecutive zero bytes L is greater than or equal to the preset threshold n, the difference R = Ln is calculated. The first n zero bytes are retained, and encoding starts from the (n+1)th zero byte. If R is less than or equal to 127, one byte is used for encoding, with the highest bit of the byte being 0 and the lower seven bits used to write the difference R. If R is greater than or equal to 128, multiple bytes are used for encoding. In the first byte of the multiple bytes, the highest bit is 1 and the lower seven bits are used to write the number of bytes required to store the difference R. The remaining bytes of the multiple bytes are used to write the difference R.

3. The method for ultra-lightweight compression of electricity meter metering data as described in claim 1, characterized in that, The process of compressing zero-byte segments in a horizontally zero-dense sequence includes the following: Set a shrinkage threshold, read bytes sequentially from the horizontal zero-dense sequence, and trigger shrinkage if the length of the zero-byte segment is greater than or equal to the shrinkage threshold. At this time, the zero-byte segment is removed, and the starting position and length of the removed segment are recorded in the shrinkage table.

4. The method for ultra-lightweight compression of electricity meter metering data as described in claim 1, characterized in that, The process of performing zero-dense transpose on the vertical zero-dense sequence includes the following: First, taking the data units in the vertical zero-dense sequence as the granularity, each byte is expanded bit by bit, and then split into 8 independent bits from the most significant bit to the least significant bit. Then, they are re-concatenated according to the bit depth priority principle to generate a reconstructed continuous data stream. Alternatively, the vertical zero-dense sequence can be converted into an N×M matrix, where N represents the number of time points and M represents the number of bytes contained in the data unit of the electrical quantity metering data. Then, the matrix can be expanded in column-major order to generate a continuous data stream.

5. The method for ultra-lightweight compression of electricity meter metering data as described in claim 1, characterized in that, In the process of residual encoding of the original measurement data sequence, if voltage data is collected, the first-order difference method is used to calculate the residual value. If power data or current data is collected, any one of the first-order difference method, two-point linear extrapolation method, and three-point median filtering prediction method is used to calculate the residual value.

6. A lightweight compression system for electricity meter metering data, employing the lightweight compression method for electricity meter metering data as described in any one of claims 1 to 5, characterized in that, include: The data acquisition module is used to collect the raw metering data sequence from the electricity meter; The residual coding module is used to perform residual coding on the original measurement data sequence to generate a vertical zero-dense sequence; The zero-dense transpose module is used to perform zero-dense transpose on the vertical zero-dense sequence to convert the vertical zero-dense sequence into a horizontal zero-dense sequence. The compression module is used to compress zero-byte segments in the horizontal zero-dense sequence to obtain a compressed sequence.

7. An electronic device, characterized in that, The method includes a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the method as described in any one of claims 1 to 5 by calling the computer program stored in the memory.

8. A computer-readable storage medium for storing a computer program that performs extremely lightweight compression on electricity meter measurement data, characterized in that, The computer program, when run on a computer, performs the steps of the method as described in any one of claims 1 to 5.

Images