3. Display units must cause interrupts in the computer only when a new particular drawing is requested by operator
3.6.2 Computer Operating System
3.6.2.1 Tasks to be performed by the computer system with the information from the process
It is now supposed that unambiguous information about the actual state of a particular component /the "State-word"/
and information about its procedence /the "Identifier-word"/
arrives simultaneously at the Interface.
162
-By means of this coded information from each component, the computer system, in the most general form, must carry out the following tasks:
- to identify the component which sends it, - to determine to which drawing it belongs,
- to determine the correct position of the symbol which represents the component within the previously selected drawing,
- to select the program /or programs/ which handles the in
formation received for updating the corresponding file /or files/ related to the drawing /or drawings/ to which the component belongs. In case the updated file refers to a drawing being shown by some display unit, the system must also update the refresh memory of the corresponding display,
- to select the program /or programs/ which handles the dynamics of drawings, in order to effectuate the visual changes within them as a product of those changes which have taken place in the state of the component.
3.6.2.1.1 Identification of the component
The information for the identification of the component must arrive at the computer from each supervised component in the
"Identifier-word" arising from the process. The total number of bits "m*" of the "Identifier-word" depends either on:
- the total number of components "N" to be supervised,
- the format given to it, which depend in turn either on the number of components "N", or on the total number of
drawings prepared for the system and the number of compo
nents per drawing, or on the number of different types of components "T" considered in the system and the number of components of each type, etc. In every case, a given num
ber of spare bits to be used in possible future expan
163
-sions or other purposes must be considered.
The particular* selection of the most appropriate format should depend on the specific characteristics of the in
dustrial plant to be supervised, although the total number of bits to have in the Identifier-word is nearly the same for all cases, because it depends finally on the total num
ber of components "N" to be supervised in the plant.
T y g e s o f Identifier-word formats
If "N" components are considered to be supervised, the fol
lowing formats could be possible:
o 1 2________________________ iog2M-i iog?N
1. ...
where: log2N is the logarithm to the base 2 of the number of components "N" to be supervised or, if the logarithm is fractional, the next larger integer.
This alternative presents the following advantages:
- It permits us to give a sequential binary order number to the code independently of the type of component, of the location in the plant, of the drawing to which it belongs, etc. It is possible then to save bits in the Identifier- word, because in such a format the minimal number are required.
- Reservation of spare bits for possible future expansions in the system or for other purposes can be done in a very straightforward form. In this case, for a large number of components "N" /in whose case "log^N" is also large/, the addition of only one bit to the total number will be
sufficient because of its considerable binary weight.
— 164 —
Since the physical location of each component in the plant is unique, the code used with this format can be easily associated to each one, thus facilitating maintainance.
It is possible to use it as an addressing-word for the computer memory# In this case indirect addressing is requi
red in order to get previously stored information about the characteristics, type, etc. of the component sending its actual "state" information.
It permits us to verify more easily the validity of the Identifier-word received.
If some expansion takes place in the system, it is not necessary to modify sustantially the programs, lists,
tables, etc. In this case, what is required is to complete the lists, tables and so on, with new data corresponding to each component added to the system.
It permits us to use consecutively the locations of the computer memory destined to store all data relative to the components /in their corresponding Component Vector/ of all drawing of the system. Thus, a relative saving of com
puter locations can be obtained, because it is not nec
essary to leave free spaces for the information related to components not existing yet.
It does not require us to store the information related to each component sequentially into the computer memory for each independent drawing or for the whole system if a
Random Access Memory /RAM/ is used. Thus, it permits us to carry out easily the modifications /enlarging/ of the
lists, tables, etc. when expansions in the plant take place.
It permits us to use a constant-length code for all the
165 -components in the system.
- Since a RAM can be used, the time required for fetching all data related to any component in the system is always the same. Besides this, it requires a relatively easy algorithm, the same for all the components.
- In case a component belongs to more than one drawing, it can be indicated by means of only one bit in the Component Vector. In this case, all data related to a given common
component which refer to every drawing of which it is a part, must be stored one after the other in the computer memory.
Disadvantages of this alternative are:
- It requires a higher number of bits per component in the computer memory to define all the data which identify it /type of component, order number within each type, etc./, because the Identifier-word is not used as data, but only as identifier.
- It uses a little more computer time per component due to the indirect addressing required to find all. the informa
tion related to the component, i.e. for searching the corresponding Component Vector.
iog2T iog2 N]
2
.
where: ,,log2T" is the logarithm to the base 2 of the number of different types of components "T" in the system or, if the logarithm is fractional, the next larger integer.
"log2N£" is the logarithm to the base 2 of the
num166
-ber of components »»N*" of the type "i" in the
system or, if the logarithm is fractional, the next larger integer#
I0ÇJ2 D loQo c
where: ,,log2D" is the logarithm to the base 2 of the num
ber of drawings "D" in the system or, if the log
arithm is fractional, the next larger integer.
"log2C" is the logarithm to the base 2 of the num
ber of hypothetic cells "C" into which the display screen has been subdivided. It gives the position of the symbol in the particular drawing "D.^" of the system. If the logarithm is fractional, the next larger integer gives the total number of bits.