EOT crane anti sway control

[mdim]

I read many articles with ABB's ASC800 drive with inbuilt anti-sway program. In the manual of the ASC800 there is no such thing, and I couldn't find any option module for this pupose.
Can anyone who have done this share some info

Not only for ABB, but for any type of equipment able to do this.

 

[JRW]

Siemens has a product called Simocrane
You can do a search for it but I have no experience with it

 

[Peter Nachtwey]

What you are asking about is motion control people call input shaping. It is usually an add on that is available on some more sophisticated controllers.

I have done this only as an experiment. It works but there isn't that big a demand for it in our market. The algorithm requires big buffers to store data so any motion controller that doesn't have lots of memory will not be capable of implementing input shaping.

The general idea is that one can determine the natural frequency of the swing by how long the cable is extended. Part of the control signal is delay by half a cycle to cancel out the oscillation. One must take into account damping too. I have the math worked out and it works well but there are limitations. For instance the delay queue also means the load will get to the desired location one cycle late but this is usually preferable to a load that swings for many cycles.

 

[mdim]

I read a couple of articles for the ASC800. where it is described as function of the drive. I was interested if anyone had done this and how it works.
Can you use it for manually operated crane, would it be acceptable for the operator, or it is preferable for kind of positioning of the load on preset locations?

 

[shooter]

i make them with fuzzy technology, as it needs more inputs then just one.

 

[mdim]

i make them with fuzzy technology, as it needs more inputs then just one.

Do you make some fuzzy logic in the drive (not familiar with ACS800, but may be something like DeviceLogix for PowerFlex)
or you make the antisway control through the PLC?

Still need some feedback about ACS800

 

[Peter Nachtwey]

There should be nothing fuzzy about sway control. Everything can be calculated
http://www.convolve.com/videos/Convolve-No-Sway-Crane-Large.mov
I talked to Marc Singhose
http://singhose.marc.gatech.edu/courses/me6404/
back in the early 1990s about this. I figured out what they are doing.

The motion can obviously be controlled manually but the the manual inputs must be 'shaped' to eliminate the sway. What is good about this algorithm is that the operator doesn't need to be smooth. I have tested this algorithm out using simple start and stop with no ramping and the convolve input shaping will smooth out the simple starts and stops. The speed is the same but the input shaping does cause a delay.

 

[mdim]

Sure, if you make the shaping in the PLC - ok.
But can you make it in the drive, if fixed speeds are used. I mean with just providing the drive with the necessary analog data (Load, cable length, speed). Because for ABB's ACS 800 drives is described this feature (only for them). If one does the shaping in the PLC, you could call the "anti-sway" control "standard feature" for any kind or series of drives

 

[BobB]

I can remember years ago seeing a demo of crane swing control using an Omron PLC fuzzy logic card. I have only ever seen it once. It certainly appeared although the crane was not huge. Would have liked to see it on a bigger crane such as a container crane at the docks where crane swing can be a pain.

 

[chud]

Out-take from a atv71 with a crane card attached to drive. Pm me if you want complete manual. I know of this being implemented at harbors here.

 

[shooter]

impossible to make it in the drive, as it needs more then one input.
so yes i make it in the PLC (Codesys)
bob the size is no matter, the length of the cable is.

 

[chud]

Can be done without plc. But networked drives are required as shown in picture. The crane card does most of the control