Hoist positioning

[CroCop]

Peter,

http://www.innocrane.com/

They have a MVI module that plugs into ControlLogix.

We originally had a SLC 5/05 installed, with HSCE2 cards. I just used the spreadsheet equations to run it up & down to test my clamping system.

We will be installing the ControlLogix this week, and should have Innocrane on site to install & train us on use of the system by hopefully the end of the month.

 

[CroCop]

You'll have to scale the Velocity to a raw units for your output card. That's pretty easy, but for my calculations and displays on HMIs, I always use engineering units, and not the raws. Raws are only for internal use in the programs I do.

 

[lauu]

Hej All

elevmike

I think I haven't been fair to that sailsman, there where more to it thant just the Encoders particulary for the Y-akse.
The measuring unit on the Y-akse consist of a mecanical divice that look like a fishingweel.
The "fishingweel" was locatet on the hoist and from here a wire is put on the actual lifting mecanismen, so when the lifting goes up/down
the wheel spins and the encoder measure the wheel spinning. And that unit is expencive to repair.

The "Fun Part" as you says is not for me to do. The "Which box goes where" is handlet by a warehouse control system made in some .net app..
The warehouse system will just tell me to pick a box at location 3005 with hoist number 3, and from that location I can calculate the
row and columen for the positioning(The function Spec. is not yet finished).

Peter Nachtwey

There will not bee any load swinging because the hoist is like a big fork lift truck.

CroCop

I think I'm on the right track now, if I understand it right the equations is only used for the UP/DOWN Ramping not for the actual
positioning.

If I'm wrong have do you now when to decelerate to reache the SP.

Jesper

 

[CroCop]

Hej All

elevmike

I think I haven't been fair to that sailsman, there where more to it thant just the Encoders particulary for the Y-akse.
The measuring unit on the Y-akse consist of a mecanical divice that look like a fishingweel.
The "fishingweel" was locatet on the hoist and from here a wire is put on the actual lifting mecanismen, so when the lifting goes up/down
the wheel spins and the encoder measure the wheel spinning. And that unit is expencive to repair.

The "Fun Part" as you says is not for me to do. The "Which box goes where" is handlet by a warehouse control system made in some .net app..
The warehouse system will just tell me to pick a box at location 3005 with hoist number 3, and from that location I can calculate the
row and columen for the positioning(The function Spec. is not yet finished).

Peter Nachtwey

There will not bee any load swinging because the hoist is like a big fork lift truck.

CroCop

I think I'm on the right track now, if I understand it right the equations is only used for the UP/DOWN Ramping not for the actual
positioning.

If I'm wrong have do you now when to decelerate to reache the SP.

Jesper

Jesper,

I know the end position before I start moving. I subtract or add (depending on position) the Dd from the End Position. That's where I start to decelerate.

 

[Peter Nachtwey]

Crocop, where did you get that those ramping equations

Your ramping equations just don't pop out of thin air.

I usually ramp as a function of time. This way I can synchronize axes. However, your equations are simple and yet still tries to follow a s-curve. I haven't had the time to break the equation down into two equations, one for position and the other for velocity as a function of time but I will do so. I am just more familiar with seeing equations as a function of time.

 

[CroCop]

Peter,

I'm taking Physics 221 (Calc Based Physics 1st Semester).

I ran into the problem of needing to accel and decel simply, and still try to follow the S-Curve. I approached my Professor and for a bargin price he came out on site and showed me a general fourth order equation. I really had to do some different stuff with the SLC 5/05 to get the math to work.

All we had to work with were Velocity on the Drives and the Encoder Positions, so we made due. It works surprisingly very well. We've run over 1000 motions over the last week, of about 20 feet total. We've set down the pipe very gently each time, so gently that every time we have corporate bigwigs touring through (we were just bought out) they show them and they are impressed.

The good Prof did it all by hand, no calculator, no math lab, nothing. Easiest money he ever made, and I was extremely impressed. I'm thinking of changing my major from EE to a Physics degree.

I need to fix the spreadsheet, I just had the copy I have at home. The one I have at work is the correct function that resides in the SLC at work. I'll post it tomorrow if I have time. (The one I posted has the general idea, but a couple of details need to be worked out).

 

[lauu]

Hej Crocop

Of course why didn't I see that (don't answer)!

I'm looking forward to see your "correct function".
I think I will try to go your way.

Thanks again
Jesper

 

[Peter Nachtwey]

Crocop, good show.

Peter,

I'm taking Physics 221 (Calc Based Physics 1st Semester).

I ran into the problem of needing to accel and decel simply, and still try to follow the S-Curve. I approached my Professor and for a bargin price he came out on site and showed me a general fourth order equation. I really had to do some different stuff with the SLC 5/05 to get the math to work.

Graduate students are also a good source of information.

All we had to work with were Velocity on the Drives and the Encoder Positions, so we made due. It works surprisingly very well.

I believe it. Most people programming PLCs do not use ramps like that. Simple linear ramps are not near as good as a s-curve.

The good Prof did it all by hand, no calculator, no math lab, nothing. Easiest money he ever made, and I was extremely impressed.

Next time give a poor grad student a chance. They may be cheaper. It may cost a 6 pack.

I'm thinking of changing my major from EE to a Physics degree.

I have been saying that most engineering is just a subset of physics for years. Mechatronics is just widening or merging Electrical, Mechanical and Computer engineering field, but physics is still a super set of them all. See Maxwell's equation. This is the first step to the unified field theory. Good stuff.

I need to fix the spreadsheet, I just had the copy I have at home. The one I have at work is the correct function that resides in the SLC at work. I'll post it tomorrow if I have time. (The one I posted has the general idea, but a couple of details need to be worked out).

I plotted your/the professor's equations using Mathcad. I still need to plot it as a function of time. To do that I need to get my calculus book out again. I don't know everything but I know there is very little that hasn't been done before. The information I need is a a book or on the web some where.

Now why do students and rookies think they are the first to see a problem? I haven't figured that out either. Oh well.

 

[CroCop]

Jesper,


This function should work. You'll need to calculate a few things before you start moving. If you have any questions feel free to ask.

 

[Peter Nachtwey]

Crocop, there is something not quite right about your equation. It may appear to work and is certainly better than nothing but it isn't quite right.
When the position is half way through the ramp the velocity should be about 1/4 of the max. In your equation the velocity is 1/2 of max velocity. Your equation has higher acceleration or deceleration at the lower speeds.

Try a simple example
x=.5*a*t^2
v = a*t

This explains why I haven't seen this equation before.

 

[CroCop]

Peter,

After doing bit of reading (and limited understanding) I know I should have a fifth order curve. This does fairly well, but is not as smooth on the accel/decel. I'll work it over & post what I get.

Thanks for taking the time to look at this.

 

[lauu]

Hej Crocop

Thanks for your time and effort.

What things do I need to calculate?

Jesper

 

[CroCop]

You'll need to calculate the Starting Position, the Decel Starting Position, and track the Current Position. You'll shift from the first equation to the second decel equation when you reach the Decel Starting Point.

 

[seppoalanen]

Hej Again

Kamenges and seppoalanen

The SQR((SP-PV) * K) + min_speed is my first choise at the moment.
Do you guys use a "Stop" sensor to stop the drive when SP=PV, or just stop when SP=PV, I'am asking because of the min_speed?
.....
Thanks to all
Jesper

Stop When SP is near the PV as "in window" +/- something.
You can play with constant K and add or sub (move curve in location-axis) some other constant, so stop happened as in servo-drive witout S-curve who makes stopping slowly.

br. Seppo

 

[lauu]

Hej Crocop

I will always now the starting point and the current point, it comes from the laser distance measuring sensor.

After looking at your spreadsheet I find this relationships: IF Dr > Da => negative result, IF Dr > Dd THEN negative result.
So I will have to do this: Acc: IF Vc > Vh OR Dr > Da THEN Vc = VH Decc: IF VC > Vh OR DR > Dd THEN Vc = Vh. Am I right?

But I can't find out how to calculate the right Da and Dr.

Hej Seppoalanen

I think I will use that stop sensor anyway, or use it as a "position control check".

Jesper