12.1. Variable Process

Under the phrase "process variable" refers to the numerical parameters defining the current state of the process. For process-related variables include the signals input / output parameters of the functional blocks, local and global flags (variables), tags, SCADA, etc.

Process variables are divided into discrete and analog. A discrete variable can take a finite number of values from a fairly narrow range. In practice, a discrete variable usually involve the value of a Boolean type (binary), indicating one of two possible states of the object (or control signal), although, formally speaking, it is not entirely correct. In the general case of a discrete variable is similar to the type of enumeration language C.

Analog variable can take any value from the limited continuous range. By type representation of the analog variable is the real number.

How Process variables are written to the archive? There are two techniques of their registration in the archive:

1. Cyclical record (cyclic archiving) implies a periodic record of the current value of process variable via user-defined intervals, regardless of the value and rate of change of this variable (see Fig. 1 - 54). Although this technique is not very economical, it is often used for archiving of analog variables. Frequency loop recording for each variable can be adjusted individually and as a rule, lies in the range from 0.5 to 10 min. As for discrete variables, and the rapidly changing analog variables, such an approach to the archive records clearly not optimal.

Fig. 1 - 54. Cyclical recording of variable process in the archive

2. Backing to change the variable (delta-archiving). This approach involves recording the variable in the archive only when a change in its value, compared with the previous recorded (absolute difference), reaches a certain value - delta (Fig. 1 - 55). The value of the delta set by the user and can be expressed in absolute units, and as a percentage of the scale. Of course, this technique is more economical than the cyclical record, as it adapts to the rate of change of archived values. For discrete variables - this approach is indispensable. Suppose we have a discrete variable, which varies, say, once per hour. Why did it back up every second or minute? It is much more logical to write down the value of the archive only in those moments when the value goes from 1 to 0 or vice versa.

Fig. 1 - 55. Delta-archiving by process variable

Where recorded archive of process variables? The most commonly used one of three options:

1. Archive written in a plain text file in format CSV (comma separated values). This file can be stored both locally and on a networked drive. Actually archive consists of a set of sequences generated files: the system generates a new archive file each shift or day. In this presentation the archive is a distinct advantage - it can be viewed using any text editor. It can also be exported to MS Excel and look at a table, applying the necessary sorting and filters. A significant drawback - it is not economical storage. Experience thus archive takes up very much space on your hard disk. To reduce the volume of the archive can be applied using compression programs, such as, ZIP or RAR - the good, text files compress very well.

2. Archive is a binary file format of which depends on the visualization process (SCADA). Obviously, it is more economical presentation of the archive, but to work with the regular program MS Excel is not enough. This archive format from different manufacturers SCADA can vary greatly. As in the previous case, the archive consists of a series created files. Generally, store video in one big file - it is not very good in terms of speed of data access.

3. The most progressive way - is keeping the archive in the form of a relational database with support for database SQL. This method allows to achieve a sufficiently high speed of work with the archive (adding records, reading and processing data). This SQL Server can provide optimal access to stories to several tens of remote clients. Since access to the archive is carried out on the open interface, SQL, developers can create custom applications to fit your needs. But the main advantage is that the archive is based on SQL - it is an excellent opportunity for the integration of automated process control system with information systems at higher levels (eg, the level of MES). As a rule, to maintain an archive SQL and customer service is used quite a powerful server platform.

In all cases described, the system archiving of process variables - is an integral part of the visualization process. The difference lies in the presentation of the archive and access technology.

What means are used to display the archive? Archives can be displayed in several ways. The easiest - is to present it in tabular format and exported, for example, in Excel, where you can draw graphs, charts, and generate reports. However, this is tedious and requires a lot of manual labor.

More convenient way - is displaying history in the form of a special dynamic (automatically updated) Graphics, called the trend (trend). Trend is placed on the operator interface to mimic those places where it is necessary and convenient to the operator. An example of the trend shown in Fig. 1 - 56.

Fig. 1 - 56. An example of the historical trend displays two Process Variables

On trend can display up to 16 variables simultaneously, as discrete and analog. This trend can be constructed for an arbitrary period of time (time span). Also supported scaling (scaling). Moving the slider (slider) along the timeline, you can view the exact values of variables at different times in the past. Segments of time during which the accident observed variables, are allocated to the trend of contrasting colors. In general, the trends - is a powerful and very handy tool, clearly demonstrates the behavior of variables in the dynamics.