JP2007221280A - Data compression method and uncompression method, and state change detection method in plant facility - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a data compression apparatus capable of efficiently compressing data and completely uncompressing the data without causing a heavy load. <P>SOLUTION: The data compression apparatus is provided with: a differential processing section 44 for obtaining a difference value between data adjacent to each other in a data sequence arranged in time series; a coding type discrimination section 45 for comparing the difference value with a comparison value and discriminating whether dynamic coding or static coding is to be applied to the difference value; a dynamic coding section 48 for applying the dynamic coding to the difference value when the difference value is smaller than the comparison value; a static coding section 49 for applying the static coding to the difference value when the difference value is the comparison value or over; a header information generating section 50 for inputting the position of the data subjected to the static coding, the length of the dynamic coded data sequence, and the length of the static coded data sequence to produce header information; and a compression data integrating section 51 for integrating the header information, the dynamic coded data sequence, and the static coded data sequence. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a data compression method, a data restoration method, a state change detection method in a plant facility, and an apparatus for carrying out these methods.

As a technique for efficiently compressing data, there is a method in which a certain number of data is collected and converted into a pseudo image and then compressed by irreversible image compression processing (see, for example, Patent Document 1).
Further, there is dynamic coding as a technique that increases the processing load even though the compression ratio is high. This dynamic coding is a method of dynamically changing the code to be assigned according to the data generation probability. Examples of this category include a Huffman code and an arithmetic code.

On the other hand, since the processing rate is low although the compression rate is low, static coding is a technique that can be processed at high speed.
This static encoding is a method of assigning a fixed code according to the occurrence probability of data, and this category includes, for example, a positive integer universal expression.
JP 2001-320278 A

  By the way, when a large number of control devices are controlled by a control device, such as plant equipment, and data is collected from measurement devices such as sensors provided to monitor the process, a lot of data is transferred on the transmission path. Will flow.

  Then, these data are collected and analyzed in the plant equipment, and operation information in the plant (of course, information on the process state is also included) is provided to the manager who monitors and maintains the data.

  By the way, it is not efficient for all managers to acquire skills for troubles that rarely occur and monitoring during special operation, etc. Therefore, in order to improve the efficiency in terms of the number of managers, Attempts have been made to centrally manage monitoring data collected at each facility by transferring it to a remote monitoring center. From the viewpoint of the line use cost, it is necessary to transmit sufficiently compressed data for the communication line between the monitoring center and the monitoring device in the plant facility. On the other hand, since a subtle change in the sensor leads to an abnormality analysis, it is necessary to use a compression technique that can completely restore the original data from the compressed data.

  However, since the technique disclosed in Patent Document 2 uses irreversible image compression processing, even if the data feature can be restored, it cannot be completely restored. Such a technique cannot be used.

  In addition, when using dynamic coding, the communication line with the monitoring center can be used efficiently, but for a large number of sensors such as process monitoring, the data compression device is heavily loaded and the amount of data to be transmitted Therefore, the compression apparatus that performs encoding becomes expensive.

  In addition, when using static coding, the compression device needs only a small load, but the compression rate is low. Therefore, when the communication line is shared with other plant equipment, the communication line with the monitoring center can be used efficiently. There is a problem that you can not. Of course, a load equivalent to that at the time of compression also occurs in the data restoration apparatus.

  Therefore, the present invention provides a data compression method and its restoration method capable of efficiently compressing data with a small load compared to the case where all are performed by a dynamic encoding technique, and a state change detection method in a plant facility. For the purpose.

In order to solve the above-described problem, the data compression method of the present invention includes a differentiating step for obtaining a difference value between adjacent data in a data sequence arranged in time series,
A code that compares the difference value obtained in this difference step with a preset value and determines whether the difference value is compressed by dynamic coding or static coding A type determination step,
A dynamic encoding step in which when the difference value is determined to be smaller than the set value in the encoding type determination step, the difference value is input and compression is performed by dynamic encoding;
A static encoding step of inputting the difference value and compressing by static encoding when the difference value is determined to be greater than or equal to a set value in the encoding type determination step;
A header information creating step for creating header information by inputting at least the length of the dynamic encoded data sequence and the length of the static encoded data sequence from the dynamic encoding step and the static encoding step;
The header information created in this header information creation step, the dynamic coded data sequence created in the dynamic coding step, and the static coded data sequence created in the static coding step are integrated. And a compressed data integration step of creating a compressed data string.

Further, another data compression method of the present invention is the above compression method, wherein when time-series data consisting of measurement data measured by the measurement device and control data for the control device is input, the time-series data is controlled. If it is data, it has a data type determination step passed to either the dynamic encoding step or the static encoding step,
In the header information creation step, specific information indicating which measurement device or control device is included in the measurement data or control data is included in the header information.

Further, the data compression apparatus of the present invention includes a difference processing unit for obtaining a difference value between adjacent data in a data sequence arranged in time series,
A code for comparing the difference value obtained by the difference processing unit with a preset setting value to determine whether the difference value is compressed by dynamic coding or static coding A conversion type determination unit;
A dynamic encoding unit that inputs the difference value and performs compression by dynamic encoding when the encoding type determination unit determines that the difference value is smaller than the set value;
A static encoding unit for inputting the difference value and compressing by static encoding when the encoding type determination unit determines that the difference value is equal to or greater than a set value;
From the dynamic encoding unit and the static encoding unit, a header information creation unit that creates header information by inputting at least the length of the dynamic encoded data sequence and the length of the static encoded data sequence;
The header information created by the header information creation unit, the dynamic coded data sequence created by the dynamic coding unit, and the static coded data sequence created by the static coding unit are integrated. And a compressed data integration unit.

The data restoration method of the present invention is a method for restoring a data string compressed by the data compression method,
A header information reading step of reading the length of the dynamic encoded data sequence and the length of the static encoded data sequence from the header information of the compressed data sequence;
A data extraction step for extracting dynamic encoded data and static encoded data based on the length of the data string read in the header information reading step;
A dynamic data decoding step and a static data decoding step for inputting and decoding the dynamic encoded data and the static encoded data extracted in the data extraction step;
Input the dynamic data sequence decoded in the dynamic data decoding step and the static data sequence decoded in the static data decoding step, and the difference between the static data sequence and the dynamic data sequence A decryption data integration step of integrating values to obtain difference value data consisting of an initial value and a difference value;
The method includes a data restoration step of inputting the difference value data string obtained in the decoded data integration step and restoring the original data string by sequentially adding each difference value to the initial value.

The data restoration apparatus of the present invention is an apparatus for restoring a data string compressed by the data compression method,
A header information reading unit that reads the length of the dynamic encoded data sequence and the length of the static encoded data sequence from the header information of the compressed data sequence;
A data extraction unit for extracting dynamic encoded data and static encoded data based on the length of the data string read by the header information reading unit;
A dynamic data decoding unit that inputs and decodes the dynamic encoded data extracted by the data extraction unit;
A static data decoding unit that inputs and decodes the static encoded data extracted by the data extraction unit;
The dynamic data sequence decoded by the dynamic data decoding unit is input, the static data sequence decoded by the static data decoding unit is input, and the difference value of the static data sequence is input to the dynamic data sequence. A decoding data integration unit that obtains difference value data composed of an initial value and a difference value by integrating
A data restoration unit that restores the original data sequence by inputting the difference value data sequence obtained by the decoded data integration unit and sequentially adding each difference value to the initial value is provided.

Further, the state change detection method in the plant equipment of the present invention is based on the length of the compressed data string including the event data in the header information of the data string compressed by the data compression method. A method of detecting,
The compressed data string is obtained by compressing time series data obtained by a measuring device provided in the plant facility.
A data length extraction step for extracting the length of the dynamic encoded data sequence and the length of the static encoded data sequence in the header information;
A length ratio calculating step for inputting the length of the dynamic encoded data sequence and the length of the static encoded data sequence extracted in this data length extracting step and obtaining the ratio of the lengths of each other;
A state determination step of determining the degree of change of the data by inputting the length ratio obtained in the length ratio calculation step and comparing it with a preset value obtained in advance.

Further, in the state change detection method in another plant facility of the present invention, in the state determination step of the state change detection method, the length ratio of data from one measurement device among a plurality of measurement devices related to each other is a set value. If the change exceeds the limit, it is determined that an abnormality has occurred in the measurement instrument,
If the length ratio of each data from two or more measuring devices among a plurality of related measuring devices exceeds the set value and indicates a change, the operation status of the plant equipment becomes unstable. It is a method to judge that.

Furthermore, the state change detection apparatus in the plant facility of the present invention is based on the length of the compressed data string in which event data is included in the header information of the data string compressed by the above-described data compression method. A device for detecting changes,
A data length extraction unit for extracting the length of the dynamic encoded data sequence and the length of the static encoded data sequence in the header information;
A length ratio calculation unit that inputs the length of the dynamic encoded data sequence and the length of the static encoded data sequence extracted by the data length extraction unit and obtains the ratio of the lengths of each, and
A state determination unit that inputs the length ratio obtained by the length ratio calculation unit and compares it with a preset value obtained in advance to determine the degree of change of the data.

Further, in the state change detection device in another plant facility of the present invention, the state determination unit of the state change detection device has a length ratio of data from one measurement device among a plurality of measurement devices related to each other as a set value. If the change exceeds the limit, it is determined that an abnormality has occurred in the measurement instrument,
If the length ratio of each data from two or more measuring devices among a plurality of related measuring devices exceeds the set value and indicates a change, the operation status of the plant equipment becomes unstable. Judging that it is.

  According to the above-described data compression method and compression apparatus configuration, when the difference value between adjacent data in the data string is smaller than the set value, compression based on dynamic coding with good compression efficiency is performed and the difference value is set. For large data that is greater than or equal to the value, compression based on static coding is performed, so by setting the setting value to an appropriate value, all are processed with dynamic coding with a large compression load. In comparison, it is possible to reduce the load of the compression process, and it is possible to make everything shorter than the length of the data string when compressed by static coding, and thus increase the utilization efficiency of the communication line. Can do.

In addition, according to the configuration of the data restoration method and restoration apparatus, the data string compressed by the data compression method and compression apparatus can be restored to the original data.
Further, according to the configuration of the state change detection method and the state change detection device in the plant equipment, whether or not an abnormality has occurred in the measuring device provided in the equipment from the length of the compressed data string at the time of the event occurrence Alternatively, it is possible to easily detect a state change such as whether or not the operation state of the facility is unstable.

[Embodiment]
Hereinafter, a data compression method and compression apparatus, a data restoration method and restoration apparatus, and a state change detection method and state change detection apparatus in a plant facility according to embodiments of the present invention will be described.

First, data compression and decompression will be described.
Usually, a plant facility is provided with at least a control device and a monitoring center for controlling a control device, and a plurality of plant facilities are provided in a place different from these plant facilities. Some monitoring centers perform centralized monitoring, and in this embodiment, the monitoring center that centrally monitors such plant facilities can be obtained with measuring equipment provided in the control equipment of each plant facility. It is assumed that the data is compressed in order to improve the transmission efficiency when transmitting the measured data.

Hereinafter, a monitoring system in such a plant facility will be briefly described.
That is, as shown in FIG. 1, in this monitoring system, a plurality of plant facilities 1 (only one is shown in the drawing) and these plant facilities 1 are provided at different locations (of course, there is a certain Or a monitoring center 2 that centrally monitors all plant facilities 1.

  Each of the plant facilities 1 is an example of a number of sensors (an example of a measuring device that measures measurement objects such as pressure, temperature, oxygen concentration, air supply amount, etc.) in order to grasp the operation state of the facility. The measurement device includes a device that collectively outputs data from a plurality of sensors), a control device 12 provided with 11, a control device 13 that controls a control unit of the control device 12, and this An operation chamber 14 for controlling the control device 12 by operating the control device 13 is provided, and various types of measurement data, control data, and the like among the sensors 11, the control device 13, and the operation chamber 14 are provided. A data transmission path 15 for transferring data is provided.

  In the operation room 14, a data collection unit that collects and stores measurement data from each sensor 11 provided in the control device 12 and control data issued by the control device 13 to the control unit of the control device 12. 21 and the data collected by the data collecting unit 21 are input to change the state of the plant equipment 1 [specifically, the operating state of the equipment (eg, stable, unstable, etc.) and the state of the sensor (eg, abnormal) And a data analysis unit 22 for displaying the analysis result and an operation unit 23 for outputting an operation command to the control device 13 based on the analysis result analyzed by the data analysis unit 22. Further, a data transmission unit 24 for transmitting the data accumulated in the data collection unit 21 to the monitoring center 2 that centrally monitors the data is provided.

  The monitoring center 2 stores a data receiving unit 31 that receives data transmitted from the data transmitting unit 24, and data received by the data receiving unit 31 (of course, restored data is also included). The data storage unit 32, the data received by the data receiving unit 31, and the data analysis unit 33 for analyzing the state change in the plant equipment 1 and displaying the analysis result, and the data analysis unit 33 Based on the analyzed result, an operation unit 34 that outputs an operation command to the control device 13 of the plant facility 1 and a state in which the data received by the data receiving unit 31 is input to detect a state change of the plant facility 1 A change detection device 35 is provided. The data accumulated in the data accumulation unit 32 can be input to the data analysis unit 33, and the state change can be analyzed later.

  The data compression according to the present invention is executed by the data transmission unit 24 provided in the operation room 14, and the compressed data is a data reception unit provided in the monitoring center 2. It will be restored at 31.

Hereinafter, a data compression method executed by the data transmission unit 24 and a data compression apparatus used for the compression will be described.
First, data will be described.

  As data to be transmitted, measurement data obtained by a large number of sensors 11 (for example, numerical data that changes continuously) and control data (so-called “so-called”) output to the control unit of the control device 12 at the same time. It is a control code and is data that does not change continuously but changes according to the contents of control). A unique number (unique information) is assigned to distinguish which sensor the data is from and to which control device 12. For example, when 60 sensors 11 are provided, No. 1 to No. 60 are assigned to these sensors 11, and when five control devices 12 are provided, control of these control devices 12 is performed. No.61 to No.65 are assigned to the parts.

  Generally speaking, based on FIG. 2, there are usually as many pieces of measurement data at the same time as the number of sensors 11 and also the number of control data according to the transmission destination [total is If N, N data 1 to data N exist]. Then, there are N pieces of data for each measurement time. In FIG. 2, a symbol (0) is added to indicate data at a certain measurement time, and data 1 (0) -Data N (0), and then successively increasing the number in parentheses at every predetermined measurement interval time, the data at each measurement time is expressed. For example, when data related to the n-th measurement is represented, the number in parentheses is (n-1). Since the control data is not issued every time, there are cases where the control data itself exists and empty data which does not exist.

  At the time of data transmission, the data for a predetermined period (for a predetermined measurement time interval) is rearranged in time series for each unique number so that the data can be easily compressed.

  For example, as shown in FIG. 3, measurement data and control data are rearranged in time series for each sensor 11 and control device 12. In this way, compression is performed on the data arranged for each unique number.

Hereinafter, a data compression apparatus will be described with reference to FIG.
The data compression device 41 inputs measurement data and control data for a predetermined measurement time interval, and rearranges the data by a data rearrangement unit 42 that rearranges each unique number in time series. A data type discriminating unit 43 that inputs a data string for each replaced unique number and discriminates whether it is measurement data or control data based on the unique number, and the data type discriminating unit 43 measures the measurement data When it is determined that the difference value is not the initial value, a difference value that is a difference from the previous data is obtained (differentiated). An encoding type discriminating unit 45 that discriminates the type of encoding by comparing the difference value with a preset value (hereinafter referred to as a comparison value), and a comparison value in the encoding type discriminating unit 45 Yo The difference value is input when it is determined to be small, and a mark value (for example, the comparison value itself) indicating that when the encoding type determination unit 45 determines that the difference value is equal to or greater than the comparison value. The first data string creation unit 46 for creating the encoding data by inputting the difference value when the encoding type determination unit 45 determines that the comparison value is greater than or equal to the comparison value. In order to compress the data by dynamic coding by inputting the second data string creating unit 47 for creating the coding data and the coding data created by the first data string creating unit 46 The dynamic encoding unit 48, the static encoding unit 49 for inputting the encoding data generated by the second data string generating unit 47 and compressing the data by static encoding, and the data The unique number obtained by the type discriminating unit 43 is And input the length data of each compressed data string created by the dynamic encoding unit 48 and the static encoding unit 49 and an identification code indicating the head position, a unique number, and an event code (event Header information generating unit 50 that arranges data in a predetermined order to generate header information, header information generated by the header information generating unit 50, dynamic encoding unit 48, and static encoding unit The compressed data integration unit 51 is configured to input the compressed data from 49 and integrate the data to obtain predetermined transmission data.

  In the plant equipment that starts and stops every day, the event code indicates that it is important data such as its start time, stop time, and monitoring time. For example, according to an instruction from the control device 13, A code corresponding to the meaning and contents is added. This event code is added to the case where the sensor indicates an abnormality in addition to the event described above.

  Further, in the data type determination unit 43, when the input data is control data, a predetermined encoding process is performed. For example, when the control data is determined to be statically encoded, the control data is sent to the second data string creation unit 47. In the present embodiment, description will be made assuming that static encoding is performed (of course, when control data is determined to be dynamically encoded, the control data is sent to the first data string creation unit 46).

  In the difference processing unit 44, if the value is an initial value, the difference is not performed, and the initial value is sent as it is to the first data string creation unit 46 (Note that the initial difference object is set to zero and the initial value is set. The value may be differentiated).

  Here, an example of time-series data before encoding when encoding time-series data for n times, specifically, an example of differentiated data created by the first data string creating unit 46 is shown in FIG. Shown in a).

  That is, in order, initial value, difference value 1, mark value (comparison value), difference value 3, difference value 4,..., Difference value (n-3), mark value (comparison value), difference value [ It is the difference between the n-th data and the (n−1) -th data] (n−1), and this data string is input to the dynamic encoding unit 48, where a known compression method ( For example, it is compressed by Huffman coding, arithmetic coding, etc. The time series data created by the second data string creation unit 47 at this time is shown in FIG. That is, the data corresponding to the place of the mark value in the difference data is the difference value 2 and the difference value (n−1), and this data string is input to the static encoding unit 49 ( Again, the compression is performed by a known method such as a universal representation of a positive integer).

  Further, as shown in FIG. 6, the header information is roughly divided into identification information, a length of a dynamic encoded data sequence, and a length of a static encoded data sequence. As described above, is composed of an identification code for indicating the head of data, a unique number, and an event code indicating data at the time of an event.

Then, the header information, the dynamic encoded data, and the static encoded data are input to the compressed data integration unit 51, and transmission data as shown in FIG. 7 is created.
Next, a data restoration device provided in the data receiving unit 31 provided in the monitoring center 2 will be briefly described with reference to FIG.

This data decompression device decompresses the data compressed by the data compression device 41 and restores it to the original data, and is executed by a procedure reverse to the encoding.
That is, as shown in FIG. 8, the data restoration device 61 includes a header information reading unit 62 that inputs header information and reads the contents thereof, and a dynamic encoding of the header information obtained by the header information reading unit 62 Based on the length of the data string and the length of the static encoded data string, a data extraction unit 63 for extracting the data string, and the dynamic encoded data extracted from the data extraction unit 63 are input and decoded. The dynamic data decoding unit 64 that performs (decompression), the static data decoding unit 65 that inputs the static encoded data extracted by the data extraction unit 63 and performs decoding (decompression), and the above The dynamic data sequence expanded by the dynamic data decoding unit 64 is input and the static data sequence expanded by the static data decoding unit 65 is input, and the static data sequence is input to the place of the mark value in the dynamic data sequence. Data The decoded data integration unit 66 for obtaining the difference value data including the initial value and the difference value by inputting the difference value data string integrated by the decoded data integration unit 66, and the difference value as the initial value. From the time series data for each unique number restored by the data restoration unit 67, the measurement data and control in all the sensors 11 for each measurement time are restored. The data rearrangement unit 68 rearranges the data into a data string.

  Note that the compression and decoding processes in the encoding units 48 and 49 in the data compression device 41 and the decoding units 64 and 65 in the data decompression device 61 use a lossless method, and thus the data is completely restored. Will be.

Next, the monitoring operation at the monitoring center 2 will be described.
In the plant facility 1, an operation command is output from the operation unit 23 to the control device 13, and the measurement data from each sensor 11 and the control device 13 control each control device 12 in a state where each device is operating. The control data at that time issued to the unit is collected in the data collection unit 21 of the operation room 14 via the data transmission path 15. At this time, the event codes from the control device 13 are also collected in the data collection unit 21.

  The data collected in the data collection unit 21 is input to the data analysis unit 22, where it is analyzed and the result is displayed. Based on the displayed analysis result, the worker outputs an operation command to the control device 13.

  On the other hand, the data collected in the data collection unit 21 is stored and sent to the monitoring center 2 via the data transmission unit 24. The data compression device 41 provided in the data transmission unit 24 efficiently stores the data. Compression is performed.

  That is, the measurement data from each sensor 11 and the control data from the control device 13 stored in the data collection unit 21 are input to the data rearrangement unit 42 by a predetermined number of times, and here, based on the unique number Measurement data and control data are rearranged in time series.

  Next, the data string rearranged for each unique number is input to the data type discriminating unit 43, where it is determined whether or not to perform the differentiation process based on the unique number. In short, a difference process is performed on the measurement data, but a difference process is not performed on the control data, and the data is sent to the second data string creation unit 47. When the event code is included, the event code is sent to the header information creation unit 50 together with the unique number.

  Then, the data that is to be differentiated by the data type discrimination unit 43 is input to the differentiation processing unit 44, and a difference value from the previous data is obtained. Note that, with respect to the initial value (there is no data before taking the difference in the first data), the initial value is sent to the first data string creation unit 46.

  On the other hand, with respect to the second and subsequent measurement data, a difference value from the previous data is obtained, and the difference value is compared with a preset comparison value and the magnitude thereof to perform dynamic encoding. Or whether to perform static encoding.

  When the difference value is smaller than the comparison value, it is sent to the first data string creation unit 46 for performing dynamic coding. On the other hand, when the difference value is equal to or larger than the comparison value, static coding is performed. In order to do this, the difference value is sent to the second data string creation unit 47 and the mark value (for example, comparison value) is sent to the first data string creation unit 46. Of course, this mark value is arranged at the position of the differential value to be statically encoded in the dynamic encoding data string.

Then, the data strings obtained by the data string creating units 46 and 47 are sent to the dynamic encoding unit 48 and the static encoding unit 49, and are encoded or compressed, respectively.
The length data of the data string encoded (compressed) by the encoding units 48 and 49 is sent to the header information generation unit 50, and header information in a predetermined format is generated. Here, an event code is added.

  The header information, dynamic encoded data, and static encoded data are sent to the compressed data integration unit 51, where transmission data is created and transmitted from the data transmission unit 24 to the monitoring center 2 side. .

According to the above-described procedure, the measurement data and the control data are efficiently encoded, that is, compressed, so that data transmission can be performed quickly and efficiently.
That is, each data from a plurality of plant facilities transmitted at predetermined time intervals is taken out in time series, and the difference between measured data is taken out of the data arranged in this time series, and the difference value is small. Since dynamic coding is performed for those that are large, and static coding is performed for those that have a large difference value, the load on the coding apparatus is reduced as compared with the case where all are dynamically coded. Can be reduced. Note that data whose measurement value itself hardly changes is compressed by dynamic encoding, so that it is possible to improve data encoding efficiency (compression efficiency). That is, when the operation state, that is, the control state is stable and normal, sufficiently compressed data is transmitted, but when a sudden event occurs or when control is hunted, the length of the data string is reduced. become longer.

On the other hand, the control data is sent to the static encoding unit 49 and encoded. However, when the control is stable, the frequency of emission is reduced, and the compression rate is increased.
In other words, the efficiency of the control data may not be improved by the same processing as the time-series change data. This is because the control data is unlikely to change every time it is measured, and a wasteful load occurs when such data is subjected to compression processing by taking a difference. For example, since the change occurs twice, once when the control data is not emitted and once when the state is returned from the emitted state, the data length is excessively increased. However, this situation can be prevented.

Next, the operation of restoring received data received by the monitoring center 2 will be briefly described.
That is, when transmission data from the data transmission unit 24 is received by the data reception unit 31, the data is input to the header information reading unit 62, and the content of the header information is read. Identification information such as a value, a unique number, and the length of the dynamic data string and the length of the static data string are detected.

  Next, when it is determined by the unique number that it is measurement data, each encoded data is read based on the detected length of each data string, and the dynamic data decoding unit 64 and the static data decoding The data input to the unit 65 and compressed is decoded (expanded).

  Next, the decrypted dynamic data sequence and static data sequence are input to the decrypted data integration unit 66, and the difference value of the static data sequence is returned to the place of the mark value in the dynamic data sequence, so that the initial value is initialized. Time-series difference value data including values and difference values is obtained.

Next, the difference value data is input to the data restoration unit 67, and each difference value is sequentially added to the initial value to restore the original data string.
On the other hand, when it is determined that the data is control data, the encoded data following the header information is input to the decoding unit associated with the unique number, that is, the static data decoding unit 65 based on the length of the detected data string. The compressed data is decoded (expanded), and the decoded data is input to the data rearrangement unit 68 via the decoded data integration unit 66 and the data restoration unit 67. As described above, for the data that has not been subjected to the differentiation process, the data decoded by the static data decoding unit 65 passes through the decoded data integration unit 66 and the data restoration unit 67 as it is. The data is input to the sorting unit 68. When the control data is dynamically encoded, it is naturally decoded by the dynamic data decoding unit 64, and this decoded control data is also passed through the decoded data integration unit 66 to be data. The data is input to the restoration unit 67.

  Then, the restored measurement data sequence and the decoded control data sequence are input to the data rearrangement unit 68 and rearranged from the time series data for each unique number to the data sequence for each measurement time. The control data has been described so as to pass through the decoded data integration unit 66 and the data restoration unit 67. However, the control data may be input directly from the decoding units 64 and 65 to the data rearrangement unit 68. Good.

By this process, the original data string is obtained.
According to the configuration of the data restoration method and restoration apparatus, the data string compressed by the data compression method and compression apparatus described above can be restored to the original data. If compression is performed by a lossless encoding technique, it can be completely restored.

Here, the main part of the above-described data compression method is described in a step format.
That is, this data compression method includes a difference step for obtaining a difference value between adjacent data in a data sequence arranged in time series, and
A code that compares the difference value obtained in this difference step with a preset value and determines whether the difference value is compressed by dynamic coding or static coding A type determination step,
A dynamic encoding step in which when the difference value is determined to be smaller than the set value in the encoding type determination step, the difference value is input and compression is performed by dynamic encoding;
A static encoding step of inputting the difference value and compressing by static encoding when the difference value is determined to be greater than or equal to a set value in the encoding type determination step;
A header information creating step for creating header information by inputting at least the length of the dynamic encoded data sequence and the length of the static encoded data sequence from the dynamic encoding step and the static encoding step;
The header information created in this header information creation step, the dynamic coded data sequence created in the dynamic coding step, and the static coded data sequence created in the static coding step are integrated. And a compressed data integration step for creating a compressed data string,
Further, when time-series data composed of measurement data measured by the measurement instrument and control data for the control instrument is input, if the time-series data is control data, a dynamic encoding step and a static encoding step A data type determination step to be passed to any of the above is provided before the difference processing step.

The main part of the data restoration method described above is described in a step format.
That is, the data restoration method includes a header information reading step of reading the length of the dynamic encoded data sequence and the length of the static encoded data sequence from the header information of the compressed data sequence,
A data extraction step for extracting dynamic encoded data and static encoded data based on the length of the data string read in the header information reading step;
A dynamic data decoding step and a static data decoding step for inputting and decoding the dynamic encoded data and the static encoded data extracted in the data extraction step;
Input the dynamic data sequence decoded in the dynamic data decoding step and the static data sequence decoded in the static data decoding step, and the difference between the static data sequence and the dynamic data sequence A decryption data integration step of integrating values to obtain difference value data consisting of an initial value and a difference value;
The method includes a data restoration step of inputting the difference value data string obtained in the decoded data integration step and restoring the original data string by sequentially adding each difference value to the initial value.

Next, the state change detection device 35 in the plant facility 1 will be described.
The state change detection device 35 is configured to generate dynamic encoded data when an event occurs, for example, at the time of power-on or during termination processing for power-off, or at a predetermined time (including an emergency situation). By comparing the length of the sequence with the length of the statically encoded data sequence, the state change in the plant equipment related to the data is detected. In other words, a long dynamic encoded data string indicates that there are many differences in measurement data that are smaller than the comparison value, and conversely, a long static encoded data indicates that measurement data It shows that there are many data whose difference value of data is larger than the comparison value.

  Therefore, by examining the ratio between the dynamic encoded data and the static encoded data, the state change in the plant equipment, specifically, the operation state of the equipment (for example, stable, unstable, etc.) and the state of the sensor ( For example, it is possible to know whether or not there is an abnormality.

  That is, in this state change detection device 35, as shown in FIG. 9, the header information reading unit 62 applies the data having the event code in the data string obtained by the data restoration device 61 of the data receiving unit 31. Data length extraction unit 71 for extracting the length of the dynamic encoded data sequence and the length of the static encoded data sequence in the obtained header information, and the dynamic code extracted by this data length extraction unit 71 A length ratio calculation unit 72 that inputs the length of the encoded data sequence and the length of the static encoded data sequence and calculates the ratio of the lengths of the encoded data sequence, and the length determined by the length ratio calculation unit 72 When the ratio is input and compared with a normal value (a ratio obtained in advance in the normal state, which is also a set value) and the length ratio is larger than the normal value, the degree of change in the data is large. Judgment When the length ratio is equal to or less than the normal value, the state determination unit 73 determines that the degree of change is small, and one sensor 11 related to the measurement data for which the state determination unit 73 determines that the degree of change is large (one Sensor 11 by determining whether the measurement object is related to each other (for example, pressure, temperature, oxygen concentration, etc. for the same combustion region) or a plurality of (two or more). Or the change state output part 74 which outputs that the driving | running state of an installation is unstable is comprised. Of course, the presence of the event code can also be known from the header information read by the header information reading unit 62.

  The normal value may be simply given as a numerical value in advance. Further, it may be obtained from data determined to be normal among the data at the same event in the past, and the determination of normality and abnormality in this case can be recognized from the result of decompressing and restoring the data, for example, For data that is determined to be normal, at least the event code and the length of the compressed data string may be stored, and the average of these may be obtained. You may make it obtain | require from the moving average of.

In addition, when comparing, it may be compared directly with the normal value, and when the difference from the normal value or the deviation from the normal value exceeds a predetermined value, it is determined that there is an abnormality. Also good.
As described above, the data receiving unit 31 receives the transmitted data, and the state of the length of each data string included in the header information is detected when the data restoring device 61 detects the event code. Measurement is performed by inputting to the device 35 and calculating the ratio between the length of the dynamically encoded data string and the length of the statically encoded data string and comparing the length ratio with a normal value. It is possible to easily and easily detect whether there is a change in data, that is, whether an abnormality has occurred in the sensor, or whether the operating state of the equipment is unstable.

Furthermore, the main part of the state change detection method in the plant equipment 1 mentioned above is described in a step format.
That is, this state change detection method is a method for detecting a state change in plant equipment based on the length of the compressed data string including event data in the header information of the compressed data string,
The compressed data string is obtained by compressing time series data obtained by a sensor provided in a plant facility.
A data length extraction step for extracting the length of the dynamic encoded data sequence and the length of the static encoded data sequence in the header information;
A length ratio calculating step for inputting the length of the dynamic encoded data sequence and the length of the static encoded data sequence extracted in this data length extracting step and obtaining the ratio of the lengths of each other;
A state determination step of inputting the length ratio obtained in the length ratio calculation step and comparing the preset value obtained in advance to determine the degree of change in the data;
In this state determination step, it is determined whether there is only one sensor related to the measurement data determined to have a large degree of change or a plurality of measurement objects related to each other, and the abnormality of the sensor or the operating state of the equipment And a change state output step for outputting that is unstable.

Further, the state determination step will be described in detail. When the length ratio of data from one sensor among a plurality of related sensors indicates a change exceeding a set value, an abnormality has occurred in the sensor. And
When the length ratio of each data from two or more sensors among a plurality of sensors related to each other exceeds the set value and indicates a change, the operation state of the plant equipment is unstable. To be judged.

  Since the data storage unit stores the restored data in addition to each data in the header information, the data storage unit is inspected with an event code, and the length of each data string is called to change the state. The state change detection device can be provided with a function of determining (detecting).

It is a block diagram which shows schematic structure of the data compression apparatus which concerns on embodiment of this invention, a decompression | restoration apparatus, and a state change detection apparatus. It is a figure explaining the data content for demonstrating the data compression apparatus. It is a figure explaining the data content for demonstrating the data compression apparatus. It is a block diagram which shows schematic structure of the data compression apparatus. It is a figure explaining the content of the time series data for demonstrating the data compression method. It is a figure explaining the content of the header information for demonstrating the data compression method. It is a figure explaining the content of the data of the integration state for demonstrating the data compression method. It is a block diagram which shows schematic structure of the data decompression | restoration apparatus. It is a block diagram which shows schematic structure of the state change detection apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Plant equipment 2 Monitoring center 11 Sensor 12 Control apparatus 13 Control apparatus 14 Operation room 15 Data transmission path 21 Data collection part 22 Data analysis part 24 Data transmission part 31 Data reception part 32 Data storage part 33 Data analysis part 35 State change detection apparatus 41 Data compression device 42 Data rearrangement unit 43 Data type discrimination unit 44 Difference processing unit 45 Encoding type discrimination unit 46 First data sequence creation unit 47 Second data sequence creation unit 48 Dynamic encoding unit 49 Static encoding Unit 50 header information creation unit 51 compressed data integration unit 61 data decompression device 62 header information reading unit 63 data extraction unit 64 dynamic data decoding unit 65 static data decoding unit 66 decoded data integration unit 67 data restoration unit 68 data arrangement Replacement unit 71 Data length extraction unit 72 Length ratio calculation unit 73 State determination unit 74 Change state output unit

Claims (9)

A differentiating step for obtaining a difference value between adjacent data in the data sequence arranged in time series,
A code that compares the difference value obtained in this difference step with a preset value and determines whether the difference value is compressed by dynamic coding or static coding A type determination step,
A dynamic encoding step in which when the difference value is determined to be smaller than the set value in the encoding type determination step, the difference value is input and compression is performed by dynamic encoding;
A static encoding step of inputting the difference value and compressing by static encoding when the difference value is determined to be greater than or equal to a set value in the encoding type determination step;
A header information creating step for creating header information by inputting at least the length of the dynamic encoded data sequence and the length of the static encoded data sequence from the dynamic encoding step and the static encoding step;
The header information created in this header information creation step, the dynamic coded data sequence created in the dynamic coding step, and the static coded data sequence created in the static coding step are integrated. And a compressed data integration step for creating a compressed data string. When time-series data consisting of measurement data measured by a measuring instrument and control data for a control instrument is input, if the time-series data is control data, either the dynamic encoding step or the static encoding step And having a data type determination step to pass
2. The data compression method according to claim 1, wherein, in the header information creation step, the header information includes specific information indicating which measurement device or control device is the measurement data or control data. A differentiating processing unit for obtaining a difference value between adjacent data in a data sequence arranged in time series;
A code for comparing the difference value obtained by the difference processing unit with a preset setting value to determine whether the difference value is compressed by dynamic coding or static coding A conversion type determination unit;
A dynamic encoding unit that inputs the difference value and performs compression by dynamic encoding when the encoding type determination unit determines that the difference value is smaller than the set value;
A static encoding unit for inputting the difference value and compressing by static encoding when the encoding type determination unit determines that the difference value is equal to or greater than a set value;
From the dynamic encoding unit and the static encoding unit, a header information creation unit that creates header information by inputting at least the length of the dynamic encoded data sequence and the length of the static encoded data sequence;
The header information created by the header information creation unit, the dynamic coded data sequence created by the dynamic coding unit, and the static coded data sequence created by the static coding unit are integrated. A data compression apparatus comprising: a compressed data integration unit. A method for restoring a data string compressed by the data compression method according to claim 1,
A header information reading step of reading the length of the dynamic encoded data sequence and the length of the static encoded data sequence from the header information of the compressed data sequence;
A data extraction step for extracting dynamic encoded data and static encoded data based on the length of the data string read in the header information reading step;
A dynamic data decoding step and a static data decoding step for inputting and decoding the dynamic encoded data and the static encoded data extracted in the data extraction step;
Input the dynamic data sequence decoded in the dynamic data decoding step and the static data sequence decoded in the static data decoding step, and the difference between the static data sequence and the dynamic data sequence A decryption data integration step of integrating values to obtain difference value data consisting of an initial value and a difference value;
A data restoration step for inputting the difference value data sequence obtained in the decryption data integration step and restoring the original data sequence by sequentially adding each difference value to the initial value. How to restore data. An apparatus for restoring a data string compressed by the data compression method according to claim 1,
A header information reading unit that reads the length of the dynamic encoded data sequence and the length of the static encoded data sequence from the header information of the compressed data sequence;
A data extraction unit for extracting dynamic encoded data and static encoded data based on the length of the data string read by the header information reading unit;
A dynamic data decoding unit that inputs and decodes the dynamic encoded data extracted by the data extraction unit;
A static data decoding unit that inputs and decodes the static encoded data extracted by the data extraction unit;
The dynamic data sequence decoded by the dynamic data decoding unit is input, the static data sequence decoded by the static data decoding unit is input, and the difference value of the static data sequence is input to the dynamic data sequence. A decoding data integration unit that obtains difference value data composed of an initial value and a difference value by integrating
A data restoration unit that inputs the difference value data sequence obtained by the decoded data integration unit and restores the original data sequence by sequentially adding each difference value to the initial value. Restoration device. A method for detecting a state change in plant equipment based on the length of a compressed data string including event data in header information of a data string compressed by the data compression method according to claim 1,
The compressed data string is obtained by compressing time series data obtained by a measuring device provided in the plant facility.
A data length extraction step for extracting the length of the dynamic encoded data sequence and the length of the static encoded data sequence in the header information;
A length ratio calculating step for inputting the length of the dynamic encoded data sequence and the length of the static encoded data sequence extracted in this data length extracting step and obtaining the ratio of the lengths of each other;
And a state determining step for determining the degree of change in the data by inputting the length ratio determined in the length ratio calculating step and comparing the length ratio with a preset value determined in advance. State change detection method. In the state determination step, if the length ratio of data from one measuring device among a plurality of measuring devices related to each other exceeds the set value, an abnormality has occurred in the measuring device. And
If the length ratio of each data from two or more measuring devices among a plurality of related measuring devices exceeds the set value and indicates a change, the operation status of the plant equipment becomes unstable. The state change detection method according to claim 6, wherein the state change detection method is determined. An apparatus for detecting a state change in plant equipment based on a length of a compressed data string in which event data is included in header information of a data string compressed by the data compression method according to claim 1 or 2. ,
A data length extraction unit for extracting the length of the dynamic encoded data sequence and the length of the static encoded data sequence in the header information;
A length ratio calculation unit that inputs the length of the dynamic encoded data sequence and the length of the static encoded data sequence extracted by the data length extraction unit and obtains the ratio of the lengths of each, and
In the plant equipment, comprising: a state determining unit that inputs the length ratio obtained by the length ratio calculating unit and compares the set value obtained in advance with the determined value to determine the degree of change of the data. State change detection device. In the state determination unit, if the length ratio of data from one measuring device among a plurality of measuring devices related to each other exceeds the set value, an abnormality has occurred in the measuring device. And
If the length ratio of each data from two or more measuring devices among a plurality of related measuring devices exceeds the set value and indicates a change, the operation status of the plant equipment becomes unstable. The state change detection device according to claim 8, wherein the state change detection device is determined to be.

Images