JP2003132093A - Data log system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the access speed of a database storing time series operation data, and also prepare for data destruction in the database. SOLUTION: Through layering and arranging a plurality of spatial positions using a general control computer 1, a plurality of CPUs 21-24 under the control of the general control computer 1, are made to register the plurality of spatial positions and state data into attached high capacity memories 31-34 along with temporal data. When an operation request is sent from an external CPU 5, the general control computer 1 makes the plurality of CPU 21-24 under the control thereof execute operation and report the results of the operation to the external CPU 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to management of status data collected during process operation, and more particularly to management of its database.

[0002]

2. Description of the Related Art Conventionally, for example, during operation of a blast furnace, state data at a spatial position inside the blast furnace is recorded in a data logger at regular time intervals or when the state changes. For example, the codes C2 and C representing the spatial position where the data are collected
3 and the state data collected as time series data are shown in FIG.
It is recorded as a set together with time data consisting of date and time as shown in 0. The state data includes temperature, pressure, level, etc., and is collected from a plurality of spatial positions.

Here, a plurality of spatial positions and state data are integrated together with time data to form a database,
Suppose you are recording to a data logger. Here, in order to manage the database in a tree structure, it is general to classify a plurality of spatial positions in a tree structure. For example, regarding the spatial position A1, if B1 to B3 are spatial positions belonging to the spatial position A1, and spatial positions C1 to C3 are spatial positions belonging to the spatial position B1, a tree structure is formed. It Therefore, for example, the state data of the spatial positions C1 to C3 depends on the state data of the spatial position B1. But,
It seems that the conventional data loggers focus only on the recording of data and are not configured to fully utilize the tree structure of data.

[0004]

In such a tree structure database, data of spatial positions are arranged in a hierarchical structure of A, B and C, and the order of priority is fixed. However, since the data of such a spatial position is not normally registered in the data logger in duplicate or in triple, there is no redundancy against the destruction of data, and the system redundancy is considered. From the viewpoint, it was difficult to secure sufficient reliability. In addition, traditional data loggers focus on recording data,
The utilization was not fully considered. Therefore, it seems that the original purpose of registering the state data in a format that is easy to use has not been sufficiently achieved. Further, in such a tree-structured state data registration system, even if an attempt is made to retrieve the state data by searching for a lower spatial position (for example, C1), the existing position of the data is sequentially determined from the upper spatial position. Since it is necessary to trace the information, it is difficult to easily perform a search in a short time.

[0005]

A data log system according to the present invention is a data log system for collecting process data, processing the data in response to an external request, and outputting the data. Position data, multiple state data collected from spatial positions, and time data at the time of state data collection are collected, and the spatial position data, state data, and time data are used as data categories, and the category is set to the maximum for each category. It is provided with a plurality of CPUs for organizing data of a hierarchical tree structure as an upper level, and a plurality of large-capacity memories respectively connected to the plurality of CPUs for storing the organized data.

In the above data log system, when one of a plurality of large-capacity memories crashes, the CPU corresponding to the crash transfers the collected and sorted data to another CPU designated in advance, and the designated CP.
U has a failure handling means for storing the transferred data in a large capacity memory.

In the above-mentioned data log system, a first processing means for instructing a plurality of CPUs to execute a process in response to a command from the external CPU and responding the result of the process to the external CPU; When a CPU or large-capacity memory fails, the CPU processing corresponding to the failure can
It is provided with a centralized computer having a second processing means for executing a failure countermeasure process for instructing U to substitute.

In the above data log system, the central computer can assign a function of organizing, for each category of a plurality of CPUs, data of a hierarchical tree structure with the category being the highest level to any CPU. The data stored in the assigned CPU is converted into a designated data format and added to a large capacity memory for storage.

[0009]

DETAILED DESCRIPTION OF THE INVENTION The present invention will be described in detail below with reference to the drawings. FIG. 1 is a system diagram showing an embodiment of a data logging system configured so that a plurality of CPUs can be operated in parallel under the control of an overall computer. In FIG. 1, 1 is a general computer, 2
1 to 24 are first to fourth CPUs, 31 to 34 are first CPUs corresponding to the first to fourth CPUs 21 to 24, respectively.
~ Large-capacity memory such as a fourth hard disk (HD),
Reference numeral 4 is a blast furnace system, 5 is an external CPU, and 6 to 8 are buses. It is assumed that the CPUs 21 to 24 are provided to organize spatial position data, state data, and time data into data categories, and to organize each category into data of a hierarchical tree structure with the category as the highest level. . In the present invention, the central computer 1 manages the data logging executed by each of the CPUs 21 to 24 in order to manage the state data with good reliability. Under the control of the overall computer 1, a plurality of state data is collected from each spatial position inside the blast furnace system 4 and sent to the first to fourth CPUs 21 to 24.
Here, the spatial position represents a spatial position where a sensor for collecting data is attached, and is therefore distinguished from a position representing an address. First to fourth CPUs 21 to 2
4 has a large capacity memory such as a hard disk 3
1 to 34 are connected, and the processing result of each CPU 21 to 24 and the progress of processing are stored. For example, the state data sent from the blast furnace system 4 to the first CPU 21 via the overall computer 1 under the control of the overall computer 1 is organized by the processing of the first CPU 21,
It is integrated with the spatial position, the date of data collection, the time, etc., and registered in the large capacity memory 31 as log data.

FIG. 2 is a model diagram showing the relationship between the data structure and the data format for the combination of spatial position, time and state data. In FIG. 2, the time is {t
₀ , t ₁ , t _2, ... T _m }. Here, when the time is fixed to t ₀ , the spatial position is composed of n spatial coordinates {S ₁ , S ₂ ..., S _n } and i state data { Let p ₁ , p _2, ..., P _i } be collected. For example, n = 4, m = 0, i = 5
At this time, if attention is paid to the spatial position, as an example of the data format, S ₁ -S ₂ -S ₃ -S ₄ -t ₀ -p ₁ -p _{2-
······ p 5, S 4 -S 3} -S 2 -S 1 -t 0 -p 1 -p
_{2 - ······ p 5, S 3} -S 2 -S 1 -S 4 -t 0 -p 1
-P ₂ --- ···· Combinations such as p ₅ are included. A combination of the above-mentioned state data obtained by paying attention to the spatial position is arranged by the second CPU 22 and stored in the second large capacity memory 32, for example. Next, m = 1, n = 2, i =
5, when attention is paid to the time t, as an example of the data format, t ₀ −S ₁ −S ₂ −p ₁ −p ₂ ... P ₅ , t
_{0 -S 2 -S 1 -p 1 -p} 2 ······ p 5, t 1 -S 1 -
_{S 2 -p 1 -p 2 ······ p} 5, t 1 -S 2 -S 1 -p 1
-P ₂ ... P ₅ etc. can be mentioned. The combination of the above state data obtained by paying attention to the time is, for example, the third
The third large capacity memory 33 is organized by the CPU 23 of
Stored in. In this way, different combinations of various data can be sorted by different CPUs and stored in different large-capacity memories. Therefore, data is organized by adopting different logic combinations for each of the plurality of CPUs. With this, for example, when attention is paid to the spatial position, it is possible to give different priorities to the plurality of spatial positions, and thus it is possible to perform hierarchization for the plurality of spatial positions.

According to the present invention, a plurality of spatial position data, a plurality of state data collected from spatial positions, and time data at the time of collecting state data are collected to obtain spatial position data,
The state data and time data are hierarchized as data categories. Therefore, by organizing the formats of the combinations of the above-mentioned various state data, the spatial position, and the time, three types of combination formats as shown in FIG. 3 can be obtained. The data format 1 is a data format organized by paying attention to the spatial position S, the data format 2 is a data format organized by focusing on the time t, and the format 3 is a data format organized by focusing on the state data p. Is. Each of these data can be stored in any convenient large-capacity memory in a convenient data format. For example, when one of the large-capacity memories crashes, the corresponding CPU transfers the contents including the spatial position, the state data, and the time data to another CPU designated in advance, and
The PU stores the transferred data in a large capacity memory.
Failure countermeasures against such a memory crash are provided for all of the plurality of CPUs. Figure 4 shows CP
It is explanatory drawing which shows the processing means performed in U. In FIG. 4, in each of the CPUs 21 to 24, state data is arranged inside together with spatial position and time data, sent to the large capacity memories 31 to 34, and arranged / registered for registration in the large capacity memories. In the event of a failure such as a means, a large-capacity memory crash, a failure response means for transferring and registering all contents including spatial position, status data, and time data to another large-capacity memory, and other It has the necessary execution processing means.

Next, in response to a command from the external CPU 5, the overall computer 1 causes each of the CPUs 21 to 24 under its control to execute an operation and causes the external CPU 5 to report the operation processing result as a response. As an example of calculation, at t = t ₁ ,

[Equation 1] It is assumed that the centralized computer 1 causes each of the CPUs 21 to 24 under the control to execute a series of operations described below according to a request from the external CPU 5. Here, p ₁ , p ₂ , and p ₃ are functions of the spatial position S (x, y, z), but it is assumed that they depend only on x. That is, it is assumed that p ₁ , p ₂ , and p ₃ depend on the spatial coordinate x, but not on the coordinates y and z.
Such data actually exists. Calculation result is external C
When sent to the PU 5, the external CPU 5 can graph the calculation result. calculation results of a ₁ ~a ₄ is a graph shown in FIG. 5, for example. In FIG. 5, a _{1 to} a ₄ are plotted with respect to the x standard value.

The following calculation is shown as still another example of arithmetic processing. With S = S ₁ ,

[Equation 2] Let's run. It is assumed that p ₁ and p ₃ are functions of time t. For example, the result of this calculation is shown in the plot shown in FIG. In this way, the overall computer 1 instructs the plurality of CPUs 21 to 24 to execute processing in response to a command from the external CPU 5,
A first processing means for reporting the result of the execution processing to the external CPU 5 as a response, and a failure countermeasure processing for instructing another CPU to substitute the processing of the corresponding CPU when a failure occurs in one CPU or a large capacity memory. And two processing means. FIG. 7 shows processing means for executing processing inside the centralized computer 1. As described above, the general computer 1 includes the first processing means, the second processing means, and other necessary processing means. Further, the centralized computer 1 makes it possible to assign, to each CPU, a function of organizing data of a hierarchical tree structure with the category being the highest level for each category of the plurality of CPUs 21 to 24, and The stored data is converted to the specified data format and added to the large capacity memory for storage.

If the spatial position {S} is represented by a (x, y, z) coordinate system, and if, for example, x = 2, then y and z are respectively x =
It is limited to points on the (y, z) coordinates in the (y, z) plane of 2. Next, if x = 2 and y = 3, z is x =
2, limited to points on a straight line determined by y = 3. If the value of the coordinate system is determined in this way, the spatial position {S}
The upper point (x, y, z) will be hierarchically determined. Thus, the spatial position {S} can be given in the (x, y, z) coordinate system, but can also be given in the (r, θ, z) coordinate system. FIG. 8 shows a combination of spatial positions {S} and state data {p} in a hierarchical manner.
The spatial position {S} is (r, θ, z) as shown in FIG.
It is given in the coordinate system.

In the present invention, the system reliability is improved by providing the system with redundancy. The first feature of the present invention is that devices such as a CPU and a memory are tripled or quadrupled to improve system redundancy, as compared with a conventional dual system such as a dual disk system. Is. In addition, since the same data is collected in the attached memory of each CPU, the redundant hardware configured by connecting in parallel with triple or quadruple is fixed by adopting the integrated computer. The second feature is that the integrated computer does not use it for the intended use, but allows the computer to select it appropriately. With this, it is expected that the operability of the system can be improved more than expected by improving the reliability of hardware.

[0016]

According to the present invention, when registering the state data together with the spatial position and time data, a central computer is adopted, and a plurality of CPUs are controlled under the control of the central computer.
By organizing the data in multiple ways and registering the results in the large-capacity memory attached to each CPU, it becomes possible to easily retrieve the status data at the time of failure and the status data at the time of maintenance. At the same time, there is an effect that it is possible to secure data redundancy and reliability that can sufficiently cope with hardware failure and software failure of the large capacity memory. Also,
In response to a request from the external CPU, the centralized computer reports the calculation result of the state data to each CPU, which has the effect of simplifying the state monitoring and maintenance.

[Brief description of drawings]

FIG. 1 is a system diagram showing an embodiment of a data log system in which a plurality of CPUs operate in parallel under the control of a centralized computer according to the present invention.

FIG. 2 is a model diagram showing a formatted data structure.

FIG. 3 is a model diagram showing an example of a data format.

FIG. 4 is an explanatory diagram showing processing means executed and executed in a CPU.

FIG. 5: Each CP executed based on a request from an external CPU
It is an example of the calculation result of U.

FIG. 6 CPs executed based on a request from an external CPU
It is another example of the calculation result of U.

FIG. 7 is an explanatory diagram showing a processing unit that is executed and executed in the central computer.

FIG. 8 is an example of hierarchically expressing a combination of spatial position and state data.

FIG. 9 is an explanatory diagram showing a spatial position in a (r, θ, z) coordinate system.

FIG. 10 is an explanatory diagram showing a conventional method of organizing spatial position, time, and state data.

[Explanation of symbols]

1 ... Overall computer 21-24 ... CPU 31-34 ... Large-capacity memory 4 ... Blast furnace system 5 ... External CPU 6-8 ... bus

Claims (4)

[Claims] 1. A data log system for collecting process data, processing and outputting the data in response to a request from the outside, comprising: a plurality of spatial position data; and a plurality of data collected from the spatial positions. The state data of, and the time data at the time of collecting the state data are collected, and the spatial position data, the state data, and the time data are used as data categories, and each category is converted into a hierarchical tree structure data having the category as the top level. Multiple CPs to organize
A data log system comprising U and a plurality of large-capacity memories each of which is connected to the plurality of CPUs and stores the organized data. 2. The data log system according to claim 1, wherein when one of the plurality of large-capacity memories crashes, the CPU corresponding to the crash transfers the collected and organized data to another CPU designated in advance. At the same time, the designated CPU is a data log system having failure response means for storing the transferred data in a large capacity memory. 3. The data log system according to claim 1, wherein the plurality of Cs are provided in response to a command from an external CPU.
First processing means for instructing the PU to execute the process and for returning the result of the process to the external CPU; and one CPU
Alternatively, when a large-capacity memory failure occurs, C corresponding to the failure
A data log system comprising a centralized computer having a second processing means for executing a failure countermeasure process for instructing another CPU to substitute a process for a PU. 4. The data log system according to claim 3, wherein the centralized computer has an arbitrary function of organizing, for each of the categories of the plurality of CPUs, data of a hierarchical tree structure having the category as a top level. Allowed to be assigned to CPU, and assigned CPU
A data log system that converts the data stored in the memory into a specified data format and adds it to a large capacity memory for storage.