I just wanted an opinion on what is considered Real-Time Control. i.e. Can you consider some or any Windows software "Real-Time" or
"Supervisory"?

 

Yes, LabVIEW DSC toolkit is a programming based SCADA software,


Anuj
[email protected]

 

If Labview compiled code runs in a deterministic manner then it can be considered real-time. This will be the case when the code runs under
an operating system that can ensure deterministic conditions such as Labview RT (usually not a Windows O.S.). To perform in this way, Labview RT uses its own dedicated computer hardware and operating system, where interrupts are carefully managed to ensure software will
execute in real-time.

Software running under Windows can be considered real-time if it is not adversely affected by the delays from the various interrupts, and events are managed in whatever the 'real' timeframe is.

On the other hand, supervisory control provides updates of setpoints to real-time controllers and stores/displays data from these real-time
controllers. These events do not need to be handled at the pace the real-time controllers operate at, but still do need to be managed within a finite time period for the control process to operate successfully.

 

As "real time" I understand any kind of control that can process information from a system or machine faster than the speed of variation of
this information, and it depends of the type of application (i.e. you are counting bottles passing in front of a photoelectric device
connected to a PLC with a program than counts the number of bottles. While the scan cycle frequency of the PLC is shorter than the
"frequency" of the bottles passing in front of the photoelectric device, the PLC is working in Real-Time, because you can count all the
bottles).

Best regards.

Aquilino

 

> I just wanted an opinion on what is considered Real-Time Control.
If "real-time" means "guaranteed response to all external and internal events within a predefined time frame", any type of control should be real-time per se (if control was not real time, this would be very bad). If "real-time" means "fast", only a sub-set of control would be real-time. If "real-time" means faster than your eyes can see, then Real-Time Control refers to anything which responds faster than within probably 50ms.

> i.e. Can you consider some or any Windows software "Real-Time" or "Supervisory"?
Depends on your view of real time. Windows operating systems do not provide a real-time execution paradigm, but there are real time extensions. As you mentioned LabView, NI offers some real time stuff, which actually is a 486 on a PCI board, with some proprietary RTOS on it. You can compile LabWindows applications to it and this will run in a real time context, even if you reboot the computer.

peter

 

LabVIEW has an add-on module called DSC (Datalogging and Supervisory Control) that adds SCADA functions (tags, trending, database,
etc) to the language.

While Windows software can be "Supervisory" I would not consider it to be "Real-Time" as the out-of-the-box Windows OS is not
capable of deterministic behavior, i.e., a guaranteed response time.

Alan Brause

 

Wouldn't supervisory control imply _higher_ reliability than what it is supervising? If so, I would say no it is not, unless hosted on a platform that meets that requirement. A HMI perhaps but not supervisory control.

Regards

cww

 

The term "supervisory control" usually refers to a strategy for determining the setpoint(s) to which lower level controller(s) will operate. It doesn't have anything to do with reliability. It does imply a longer time horizon (you don't change setpoint nearly as often as you move a valve) and less stringent real time requirements.

Diana Bouchard

******************************************************************************
Diana C. Bouchard
Pulp and Paper Research Institute of Canada (Paprican)
Process Control Group
570 St Johns Boulevard
Pointe Claire Quebec H9R 3J9 Canada
phone: (514) 630 4100 x2376
fax: (514) 630 4120
email: [email protected]
******************************************************************************

 

Not necessarily. If the underlying control elements aren't working, the supervisory control won't do squat. On the other hand, if the supervisory control is not working, but the underlying control elements are, then usually the operator can take over with some level of degraded/manual/open loop control. My definition of supervisory control is one that can
coordinate the actions of multiple I/O variables and/or in multiple process states. Maybe it is a semantics problem.

Regards,

Lou Heavner
Emerson Process Management

 

Yes, quite frequently these terms are rather narrowly defined.
And logical implications aren't necessarily valid.

Regards

cww