Any Automatitions here?

[Rjhere]

I am currently studying automation, which involves relays, PLC programming, regulation of different processes, etc.

Just wondering if anyone else in this forum works with this or are studying it themselves?

Cheers

[Sun]

If you're looking for IRL animations, well, I have no comment.

If you mean computer animations, I have made a few, here's my most recent: http://forum.kerbalspaceprogram.com/threads/97981-A-new-series-of-Minecraft-animations-I-m-making

[Rjhere]

If you're looking for IRL animations, well, I have no comment.

If you mean computer animations, I have made a few, here's my most recent: http://forum.kerbalspaceprogram.com/threads/97981-A-new-series-of-Minecraft-animations-I-m-making

Can't help but feel you just wrote that to promote yourself... and that's largely irrelevant with animations when I wrote automation.

[Rainbowtrout]

If you're looking for IRL animations, well, I have no comment.

If you mean computer animations, I have made a few, here's my most recent: http://forum.kerbalspaceprogram.com/threads/97981-A-new-series-of-Minecraft-animations-I-m-making

The topic of the thread is automations. An animation about automation...

[Beowolf]

I am currently studying automation, which involves relays, PLC programming, regulation of different processes, etc.

Just wondering if anyone else in this forum works with this or are studying it themselves?

Cheers

Not me, but a close friend worked for an industrial conveyor company back around 1980. He was the one helping them transition from relay logic to PLCs and 8-bit microprocessors. Also he holds a patent on a type of efficient lane diverter. A competitor had designed a new one, and my friend was sent to look one over and figure out how to do it as efficiently as their product, but without violating their patent.

I worked with a specific type of relay logic for a while, for a company that refurbished and resold old relay-based telephone company central office equipment. But I wasn't there long enough to learn very much. After a few months, I moved to a job operating an IBM/360 mainframe, which is where my permanent career in IT got rolling.

[Camacha]

Not really, though working with kOS made me realize I do have a soft spot for control theory and associated branches of technology. I just love developing, testing and polishing the control logic to the point you are sure it is about as efficient as you can make it. Eats up massive amounts of time though, just making it work or making it just work is a little easier.

[steve_v]

I'm an industrial electrician, with some experience in automation / PLC programming. Mostly ladder logic on the AB PLC5 but starting to get into structured text and functional block.

I get to play with a fairly wide range, from 1960s analogue hand-wired TTL controllers to state of the... multithreaded multi-language PLCs.

But it's usually fixing someone elses code. or trying to read german blueprints and 'now, what does this jumper do?'

A KSP automation system in ladder would be very, very cool...

Also, PID loops can be a royal PITA IRL, despite being *just* simple math on paper...

Especialy when someone else's 'just works' solution is in place and the process... err, changes.

Some of the newer model based controllers are pretty interesting though.

Edited October 31, 2014 by steve_v
Spelling

[Camacha]

Also, PID loops can be a royal PITA IRL, despite being *just* simple math on paper...

Especialy when someone else's 'just works' solution is in place and the process... err, changes.

Some of the newer model based controllers are pretty interesting though.

I have been experimenting with PID versus model based controllers in KSP. Despite KSP being a fairly ideal environment with almost no noise and other complicating factors, the latter can be very, very time consuming to develop. If it works it is beatiful though, hover scripts that function perfectly and accurately, even when instantly switching planet or gravity through Hyperedit. It just works. I had a great time coming up with a solution that would work not only under any condition, but also under any aerodynamics and fuel model without resorting to making specific exceptions.

PID controllers are a little easier to get to work in my experience, but when the variables change it results can be less than ideal indeed. I first noticed this when I tried picking up a relevant amount of cargo with my PID hover script controlled craft. The unloaded craft stabilized quickly, the one with more weight tended to lithostabilize, eventually. Admittedly my tuning approach is fairly crude, but what makes PID controllers effective - their lack of process knowlegde, also makes them inherently unstable.

Any pointers to where I can find good info about the real thing? There is a fair amount of information to be found on PID, but when it comes to more modern controllers less information is available.

(Please note that I have no formal training in this area, so I might use the relevant terms incorrectly.)

Edited October 31, 2014 by Camacha

[Tommygun]

I took a semester of PLC programing in college and have worked a little bit with them in a factory setting.

Mostly through a graphical ladder logic interface.

If anyone is interested about learning to program a PLC, here is a free simulator:

http://www.thelearningpit.com/plc/psim/psim.html

I used it while taking the class and it's kinda fun.

[steve_v]

There is a fair amount of information to be found on PID, but when it comes to more modern controllers less information is available.

Yeah, tell me about it.

Even the vendors paid support really just says 'use our black-box IMC instruction, designed for just your application' with nill information on how it actually operates internally

This approach usually works, with a little fiddling, but one learns nothing from the exercise.

Unfortunately my job desription is implementation rather than design, so on the rare occasion an in-house solution is required, I'm just shoving a process engineers math into the controller & tweaking the knobs.

I'm a bit of a perfectionist, so the whole scenario really annoys me. But hey, time is money so once it 'kinda works' it's time to move on to the next one