In-depth understanding of floating-point rounding instructions - Database & Sql Blog Articles

SMD aluminum electrolytic capacitor

The outer diameter of the probe domestic PA50-F small four-jaw gold-plated needle is 0.68. The total length of the needle is 16

Floating point numbers are also called real numbers. In STEP 7, floating point numbers are represented by decimals. I saw on the Internet that the "ROUND command of the S7-300/400 is actually 5 rounds. If you really want to round up, you can add 0.1 in advance."
To be honest, I have not paid attention to the details of this directive in the past. To this end, first consult the online help of the directive:
“ROUND (rounded to a long integer) reads the contents of the parameter IN as a floating point number and converts it to a long integer (32 bits). The result is the nearest integer (rounded to the nearest value). A floating point number is between two integers, and an even number is returned. The result is output by the parameter OUT.
We commonly refer to the “rounding to the nearest value” as rounding off. The following is the result of the simulation experiment, it seems that it is still rounded off, there is no problem of 5 rounds.
The input is 2.49999 and the output is 2 (four rounds).
The input is 2.50001 and the output is 3 (five entries).
Floating-point numbers return even numbers between two integers, which is different from our habits, which may be five or five. The simulation results are as follows:
The input is 1.5 (between 1 and 2) and the output is 2 (five in).
The input is 2.5 (between 2 and 3) and the output is 2 (five).
The official documents of Siemens (including the help files in the software) are the basis for our study. To understand them deeply, we need to do experiments.
It should be noted that floating point numbers allow up to 6 significant digits to be entered.