JPH04348226A - Running data preserving method - Google Patents

Abstract

PURPOSE:To preserve a long-term effective data in a memory device of small capacity by preserving every specific number of running data as one group every time it is collected and preserving the average value of the adjacent two data in the previous data as a representative value. CONSTITUTION:When a first period 2<n> X time has passed after the collection of the running data was started, 2<n> data is collected in a data collecting device and preserved as a group G1 in a secondary memory device. Then, when a second period 2<n> X time has passed, another 2 data is collected in the memory device and preserved as a fresh group G1. The average value of the adjacent two time series data in the previously stored group G1 is preserved as a group G2. In this manner, the number of data in the group G2 is reduced to 2<(n-1)>. If this sequence is repeated, a long-term effective data can be preserved in a memory device of small capacity.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for storing operational data suitable for processing and storing operational data of industrial plants such as boilers and incinerators using electronic computers.

0002

[Prior Art] In plant equipment such as boilers and incinerators, calculations include operating data such as temperature, pressure, and flow rate for each part of the equipment, as well as boiler efficiency, heat transfer coefficient, converted NOx, etc., calculated using the operating data. Values, etc. are recorded, and these records are important for determining the normal operation of the equipment, fault diagnosis, deterioration over time, etc.

Conventionally, in the above-mentioned plant equipment, operating data has been recorded by operators periodically reading and recording the values of meters. In addition, as shown in FIG. 3, a transmitter is attached to the measuring instrument 4 installed in the plant 3, and the signal from this transmitter is converted from analog to digital by an A/D converter 5.
A data collection device is also used that converts the data into a computer 6, stores it in a secondary storage device 7 attached to the computer, and outputs it on paper via a printer 8 at regular intervals in the form of a daily report, monthly report, etc. There is. After outputting the operation data collected by the data collection device to a printer,
The oldest data is deleted from the secondary storage device, or when the storage capacity of the secondary storage device becomes full.

0004

[Problems to be Solved by the Invention] In order to judge the status of plant equipment, such as its normal operation, failure diagnosis, and aging deterioration, it is necessary to use operating records recorded by operators or printed out on paper from data collection devices. , it is necessary to compare and judge in chronological order.

[0005] In a data collection device using an electronic computer, operating data is stored in the secondary storage of the computer, so a secondary storage with sufficient capacity is prepared to record time-series changes in operating data. By using software that describes inspection procedures, it is possible to use a computer to determine whether plant equipment is operating properly, diagnose failures, and determine whether it has deteriorated over time, without the need for human intervention. By the way, the number of types of operational data that should be recorded is about 50 points even for relatively small boilers, and more than 2000 points for commercial boilers, etc., and the time interval for recording is every 10 seconds for equipment with fast response speed, and for ordinary boilers. Even in equipment, it is common for the transmission to occur every minute.

Therefore, even if data is collected every minute, the number of data to be recorded per day is 50 (
/ minute) x 60 (minutes/hour) x 24 (hour) = 72,00
0, and for a commercial boiler, similarly calculated, 20
00×60×24=2,880,000 pieces.

[0007] Normally, one piece of data is 2 bytes and is stored in the secondary storage device, so in the above case, if it is stored for one year, each data will be 72,000 (pieces/day) x 2 (bytes/piece) x 365 (day) = 53MB, similarly 8800
00 x 2 x 365 = 640 M bytes (M bytes = 100
(10,000,000 bytes). For example, in order to judge from operational data, such as dirt on the heat transfer surface of a boiler or clogging of the flue, it is necessary to compare changes over time in data over about one year. Therefore, in order to store all the data, it is necessary to use two very large capacity
A storage device is required, and it is difficult in practice to compare operating data with an electronic computer and make such judgments.

[0008] When investigating time-series fluctuations in operating data,
If the time series to be compared is short, operational data collected at short time intervals is required. For example, in order to examine hourly fluctuations, it is necessary to compare data over a period of approximately one minute. On the other hand, by comparing long-term driving data,
When evaluating the change, the interval between operating data to be compared may be long. For example, to examine data fluctuations over a period of about one year, data with a period of about one day is sufficient. Therefore, even if the collected driving data is regularly organized, new driving data is kept densely, and data is stored in a format where older data becomes coarser, it is difficult to understand changes in driving data over time. There will be no problem.

The present invention has been made in view of the above-mentioned circumstances, and has the practical purpose of storing new operational data closely and storing past data so as to be suitable for investigating changes over time in plant operational data. The purpose of the present invention is to provide a method for storing driving data that reduces the amount of data to be stored.

0010

[Means for Solving the Problems] In order to achieve the above object, the operational data storage method of the present invention provides a method for storing operational data periodically collected in plants such as boilers and incinerators in chronological order. First collected 2 to the power of n (n is a positive integer)
After that, every time 2<n> operation data is accumulated, the operation data is saved as one group, and each group in the previously formed group is saved. A feature of the data is that one representative value is determined based on two adjacent pieces of data, and the representative value is stored.

[0011] The representative value is preferably an average value of two adjacent data. Further, it is preferable that the driving data is automatically collected and stored in a storage device, and the representative value of two adjacent pieces of data is determined by a computer.

0012

[Operation] According to the method for storing driving data of the present invention, each time 2 n driving data is stored, the driving data is stored as one group, and each data in the group formed before is stored. Since two adjacent pieces of data are combined into one representative value and saved, newer data can be stored more densely and older data more coarsely stored, allowing long-term effective data to be stored with a small storage capacity. can do.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The method for storing operational data of the present invention will be explained in detail below with reference to the drawings. Figures 1 and 2 are flowcharts showing the procedure for compressing the number of driving data, Figure 3 is a schematic diagram showing the procedure for extracting data collected at a certain time from a compressed group of driving data, and Figure 4 is a configuration diagram of the data collection device. It is.

As a device for recording operating data in a plant such as a boiler or an incinerator, a data collecting device equipped with a computer, a secondary storage device, etc. is used as described in the section of the prior art. Collect data periodically. According to the method of this embodiment, among the collected driving data, the newest data is stored more closely, and the older the data, the more coarsely the data is stored.

Regarding the data collection method, plant operation data is collected up to (2 to the n power) pieces for each type, and these (2 to the n power) data are made into one group. Here, n is a suitable integer greater than 1. If there are multiple types of data, perform the same operation for grouping each type of data. Since this operation is the same for various types of data, one type of data will be explained below.

[0016] Letting the data collection cycle be T (time),
Letting the initial time be O, every (nth power of 2) x T (time),
The data collected during that period is grouped. The groups are numbered consecutively in chronological order such as 1, 2, 3...m, and are expressed as group Gm. In other words, the newest group is always numbered 1, the group that was number 1 is numbered 2, the group that was number 2 is number 3, and so on. , the number increases by 1 each time a new group is created. Further, the number of data collected during a period less than (2 to the n power)×T is equal to or less than 2 to the n power, and this is set as group G0.

[0017] First, when the first period (2 to the n power) x T (time) after starting the collection of driving data has passed, (2 to the n power) pieces of data are collected by the data collection device. It is saved in the secondary storage device as group G1.

[0018] Then, (2 to the n power) x T for the next second period
When (time) has passed, a new (n of 2) is added to the data collection device.
) data are collected and saved as group G1. At this time, the data already saved is
The data is processed by an electronic computer, that is, the average value of two pieces of data adjacent in time series is determined, and the average value is stored in the secondary storage device as group G2. Thus, the number of data collected during the first period is reduced from 2 to the nth power to 1/2 of the 2 to the power of 2(n-1).

[0019] Also, when (2 to the n power) x T (time) of the third period has passed and new (2 to the n power) data is collected, the 2 data collected in the second period The n-th power of data is averaged to 1/2 in the same manner as described above, and the 2(n-1) power of data is stored as group G2 and collected in the first period. For the data that has been processed and halved, the average value of two adjacent pieces of data is calculated, and the number of data is reduced to 1/2 again, that is, 1/4 from the time it was collected, and is saved as group G3. . Thereafter, each time (2 to the n power) pieces of driving data are collected, they are processed in the same manner as described above, and the number of data stored up to that point is halved. This process is expressed by the following formula.

[0020] When (2 to the n power) × T time has passed, (n-1) of data 1, 2, 3, ..., 2 of group Gi
2 to the nth power, each data is Di, group Gj
If each data of is Dj,

[0021]

[Math 1]

(i, j=1, 2, . . . , 2 to the (n-1) power) is averaged over time to generate new data.

Here, there is a relationship of j=i+1. This operation reduces the number of data in group Gi to 1/2 of the original number, and similar operations are sequentially performed on other groups.

However, if there is only one data in a group, no operation is performed and only assignment to the next group is performed. Furthermore, each data element of group G0 is assigned to the corresponding element of group 1.

[0025] The total amount S of the entire group of data compressed in this way is calculated based on the compressed portion per data type.

[0026]

[Math 2]

On the other hand, if no compression is performed, the total amount S
'teeth

[0028]

[Math 3]

[0029] If n is sufficiently large, the compression ratio is

[0030]

[Mathematical 4] S'/S=2/(n+1), and if n=14, then 1 when held normally
This means that a storage capacity of about 3% is sufficient. If the value of n is increased to 14, the effect of this data compression becomes even greater.

Furthermore, if n=14 and data is collected every minute, the time required to form one group is 2.
14 = 16384 minutes = 11.37 days, and each time a new group is created, the data accumulated so far is calculated using formula 1.
It is after 11.37 x 14 = 159 days that the number of data in the first group becomes 1, and this number of days is practical for analyzing operational data. It is reasonable.

FIGS. 1 and 2 are flowcharts showing the data compression procedure described above. The array for storing operation data by group is G0 [N0], G1 [N], G2 [N/
2]...Gm-1[4], Gm[2], and GN[M] are defined. Here, N is (2 to the power of n), where n is an appropriate integer greater than 1, and M is also another appropriate integer greater than 1.

Data processing is performed as shown in FIG.
and according to the procedure shown in FIG. Process I is a part for determining whether a predetermined number of data forming one group has been accumulated. Although not shown in the figure, data is sequentially stored in the array G0. Once a predetermined number of items have been accumulated,
Process II Execute the following. In process II, the data of each group is sequentially sent for the group having the oldest data group, such as one group containing only one piece of data. The average of the two data in the previous array Gm[2] is assigned to the vacant GN[1]. The averaging process indicates a means for extracting representative data, and other methods may be used.

Processing III and processing IV each take the average of two elements of the previous array and assign it to the element of the next array. Although it is omitted in the middle, G3~Gm-
The same operation is applied to the array of 1. Processing V then G
The elements of the 0 array are assigned to each element of G1. The operations from process II to process V are executed every time N pieces of data (2 to the nth power) are stored in the G0 array.

Next, the data collected at the specified time in the past do
We will explain how to take it out.

[0036] Find the time t until the specified time, and calculate the collection period T.
From this, the number L of data collected by time t is determined.

L=t/T If L is smaller than the number N0 of data in group G0, the desired data do is in group G0. The Lth data in group G0 is the desired data.

[0038] When L exceeds N0, a new

003
9]

[Math. 5] L0 = L-N0 Define. next

[0040]

[Math 6]

Find m that satisfies the following. Target data d
o is in the mth group Gm. next

0042
]

[Math 7]

[0043] Group Gm has 2 (n-(
There are m-1)) data, and each data is 2(m-
1) It is the average of the multiplication original data, and is treated as the target data to be obtained. Next, divide L1 by 2 to the (m-1) power to obtain an integer value of the quotient. This operation is defined as ¥.

[0044]

[Math. 8]

[0045] Then, the data to be sought is in the L2th position of the mth group Gm.

The above calculations can be easily realized using an electronic computer.

As an actual example, a procedure for extracting target data from compressed data will be explained using FIG. For simplicity, it is assumed that n=3 and one group consists of eight pieces of data. In the figure, d1, d2,...
d35 shows the collected data in chronological order, one data frame 1 contains one data, and when one data frame 1 contains multiple data, there is one representative value of them. It means that. Assume that when data is stored in five groups as shown in FIG. 3, the value d16 of L=16th data is determined. Group G0
Since there are three pieces of data, from Equation 5, L0 = 16-3 = 13 Next, from Equation 6, (m-1) x 23 < 13 < m x 23 ∴m = 2 From Equation 7, L1 = 13- (2- 1) x 23 = 5 From formula 8, L2 = (5 yen (2 to the power of (2-1)) + 1 = 3 Therefore,
The 16th data d16 to be found is 3 of group G2
The th data corresponds to this. This data is 16
, 17th data d16, d17, but since there is no problem in comparing old data with representative values, this is adopted.

[0048]

According to the present invention, the driving data storage method is such that each time 2nth power of driving data is stored, the driving data is stored as one group, and each data in the group formed before is stored next to each other. In particular, the method according to the embodiment of the present invention is based on calculations using powers of 2, and the internal calculations are performed in binary, since two pieces of data that match are combined into one representative value and stored. High-speed processing is possible on electronic computers.

[Brief explanation of drawings]

FIG. 1 is a flowchart showing the procedure of an operation data storage method according to an embodiment of the present invention.

FIG. 2 is a flowchart showing the procedure following FIG. 1;

FIG. 3 is a schematic diagram showing a procedure for extracting target data from a saved data group.

FIG. 4 is a configuration diagram of a data collection device.

[Explanation of symbols]

1 Data frame 3 Plant 4 Measuring instruments 5 A/D converter 6 Electronic computer 7 Secondary storage device 8 Printer di data (i: integer) Gm Group (m: integer)

Claims (3)

[Claims] Claim 1: In an operation data storage method for chronologically storing operation data periodically collected in a plant such as a boiler or an incinerator, the first collected 2 to the power of n (n is a positive integer) Form and save one group with driving data, and then save the driving data as one group every time 2 to the n power of driving data is accumulated, and each data in the group formed before then. 2 adjacent
1. A method for storing driving data, characterized in that one representative value is determined based on data for each item, and the representative value is stored. 2. The driving data storage method according to claim 1, wherein the representative value is an average value of two adjacent data items. 3. In an operation data storage method in which operation data automatically collected periodically in plants such as boilers and incinerators is stored in a storage device in chronological order, (integer) driving data are formed into one group and saved, and then each time 2 n driving data is accumulated, the driving data is saved as one group, and the data in the previously formed group is saved. 1. A method for storing operating data, characterized in that one representative value is determined by a computer based on two adjacent pieces of data for each data set, and the representative value is stored.