OT, It takes one to know one.

But doesn't that nullify the reason for going out on the ocean?

No... a day after out in the great lakes or any big water is a nightmare in ruff waters, I never get seasick but you get exhausted after a few hours of fight just to stand up

jstolaruk said:
yeah, they are NOT cheap.

If and a big IF you can afford one I think it would be money well spent, I have about 8 years before I am going to get back out on the big water.... I will be selling my car so I can get one, now I just need to figure out how to pay for the boat :D
 
Here is another cool idea and technology

https://www.youtube.com/watch?v=n1ob816u2ec

If I ever get back to the ocean I will be saving my money to get one...

That is not the same. The input shaping is changing the target generator to take into account the vibration and cancel it out.

The wave stabilization is reacting to waves. This is common in water applications. There is a company called Pinfabb in Italy that retrofits cruise, container and other large ships using our controller. This is just a matter of keeping the gyro level.
https://www.fluidpowerworld.com/electrohydraulic-controls-improve-ship-stability/

Hopefully, in the not too distant future, that will be available on all/most servo controllers
We have not seen that much demand for it. The algorithm has a draw back. It adds one oscillation time to each motion. So if the object oscillates at 10 hz the the motion will take 100 ms longer. The extra time can be a production killer. It should only be used in single axis applications or where all the objects under control have objects that oscillate at the same frequency and in phase. The advantage this algorithm has is that it doesn't require a fancy motion generator.

We went the fancy target generator and control algorithm route.
https://deltamotion.com/peter/Videos/NF-FOA.mp4
The natural frequency is about 6.5 Hz so the period is about 0.15 seconds. This system was designed to be almost impossible to control without the proper algorithm. I attempted to tune the system using a PID with velocity and acceleration feed forwards. You can see it fails unless I reduce the gains or speeds. I then use our control algorithm and it works again.

A little history. I am pretty sure the input shaping was discovered by MIT in the late 1980s. In 1990 a MIT student came by trying to sell me a license to the technology. We didn't have the CPU power or time to spend on the technology but it opened my eyes. I knew how it was done. Years later I saw a doctoral thesis from the student that visited me. He is now a professor at GA Tech. I made a YouTube video about this
https://www.youtube.com/user/pnachtwey/search?view_as=subscriber&query=input
There are only 164 views after all this time. It is not a hot topic. If the video got a lot of views then I would be more inclined to add this to our product.

BTW, the most popular video is the one about combining a Smith Predictor with Sliding Mode Control. It has 2.1K views. It is too bad there aren't enough geek out there to like this stuff.
 
That is not the same. The input shaping is changing the target generator to take into account the vibration and cancel it out...

I know I was just pointing out that it was also very cool engineering idea.

Pinfabb does the big boy's boats... not the little toys I play on.
 
Maybe if you got in touch with your new uncle down there (the one that just picked up a $800 million check from the Mega Millions game) he could get you a bigger boat to play with, or a few....

It was a lady so maybe my new ant :) but she was also just passing through, hope she does good with it, thats a lot of money!
 
Originally posted by Peter Nachtwey:

This mathemagic is called input shaping.

The way the OJET website makes it sound the target generator operating in spectral mode simply won't put out a position command with a frequency content that will excite the load. But in practice it looks more like this would be the same thing as sway control in cranes, where the position command to the axis basically causes the driving member to "slow down" to allow the load to catch up at the end of the accel phase and then continues on with the move. Functionally they may end up in the same place but conceptually they seem different (even though they may not be).

Keith

[EDIT] I just looked at Peter's second link. Question answered.
 
Last edited:

Similar Topics

Hi, I have a 1500 that controls a station with diferents warehouses, but i also have a 1200 that controls one of those warehouses, i have been...
Replies
9
Views
210
Hello Everyone I have a Delta VFD, after it is supplied by the three phase voltage at the beginning it displays nothing, then after about 5...
Replies
1
Views
800
Just burnt up an output card because I forgot to include a relay between the output and a contactor, so the inrush shorted the output to always...
Replies
56
Views
14,471
Hello, I have a problem with Factory TalkView Studio. I have been programming loops RSlogix 5000. I have a tag called DrumSP (Real) going into...
Replies
3
Views
1,611
I'm sorry to bother you guys, I have been battling this random issue for weeks. We have 14 - Kinetix 6000 drives. When the operator sometimes...
Replies
3
Views
4,126
Back
Top Bottom