vibraiton isolation

[rahulbanik]

Hi
 
I am working on a six degree of freedom active low frequency (0.2-100.0Hz) vibration isolation system. I have to build a vibration controller based on feedback from six Seismic accelarometer Wilcoxon  731-201 and signal conditioner (IEPE)  and then provide the driving control signal to six Voice coil motor through the drivers considering the kinematics of the system and control law implemented.
I was wondering what kind of controller do I need for such six degree of freedom vibration control especially in the low frequency region i.e. around 0.2 - 100 Hz.
I am eagerly waiting to get replies from you.

Thanks

Rahul

[Jochen]

Rahul,

you should be able to tackle this application with the NI SoftMotion development module for LabVIEW. Please read this article from our developer zone to learn more about this software module. Here you can find very detailed information about all NI motion control options. You should use this link later as a reference for looking up specific details.

Depending on your requirments (power consumption, footprint, speed,...) there are multiple hardware options available as discussed in the article above. A combination of Real-Time and FPGA targets (e. g. PXI with a RIO board or a cRIO system) provide the best performance and the highest level of flexibility. The most inexpensive solution is a PC running LabVIEW Real-Time and a data acquisition board but this option is probably not powerful enough to control 6 axes in parallel.

The NI SoftMotion development module provides a lot of great features that allow an experienced LabVIEW programmer to easily build a customized motion controller.

Please let me know if you have further questions.

Best regards,
Jochen Klier
National Instruments Germany

[Brian_Beal]

If you want an off-the-shelf hardware solution, then I would recommend a Delta Tau controller.  They can be ordered with analog feedback, and can inherently do coordinate transformations.  Labview drivers are available.  Make sure that the conversion rate of your signal conditioner is fast enough for your application,

Message Edited by Brian Beal on 03-27-2007 06:57 AM

[Jochen]

Brian,

I was thinking of OTS-solutions, too but I think this application requires probably more degrees of freedom than provided by those solutions. I'm pretty sure that just a coordinate transformation won't be enough to solve this application. The alogrithm will have to take care of the kinematics (serial or parallel kinematics?) and it will have to calculate positions from the acceleration sensor signals. I don't believe that this will be possible with a standard off-the-shelf product.

By the way: A cRIO system provides direct connectivity for IEPE sensors without the need of external signal conditioning and it's possible to use customized control algorithms (e. g. model based controllers).

Jochen

[Brian_Beal]

Jochen-

Actually, the Delta Tau PMAC can use analog inputs for the algorithms.  I am almost positive that it handle 6 degrees of freedom without any problem.  It is a very powerful controller that can be configured for many different applications.  You can even write your own algorithms and have them run on the PMAC.

Brian

[rahulbanik]

Dear Brian and Jochen

Thanks a lot for your replies.

I got a NI PXI 7350 motion controller. Can you please inform me whether it can be implemented for my system? I wonder whether it can measure the low frequency  (0.2-100.Hz)  accelaration. Also in order to implement the PI algorithm and to control the accelarometer drift I need to integrate the accelaration signal to get the displacement and velocity. Can you please inform me how I can do that using the controller.

Thanks again

Rahul

[Jochen]

Rahul,

I don't think that the PXI-7350 is a good choice for your application. It's a great product for standard motion control tasks but as you can't modify the feedback signals in real-time onboard and as there is no way to use another control algorithm than the implemented PID, I don't think that this product will provide enough flexibility for your application. That's why I haven't mentioned this board.

As you are obviously using a PXI system, you could use NI-SoftMotion, NI-DAQ boards and/or NI-RIO boards as I have suggested in my first answer.

Best regards,

Jochen

Message Edited by Jochen on 04-03-2007 09:28 AM

[rahulbanik]

Thanks again for your suggestions.

I have decided to use R series NI -PCI 7831R (8ch) according to your suggestions. But considering the price I decided to use Labview FPGA module only. I can not afford Soft Motion development module due to budget constraint.

Can you please inform me whether it is possible to measure the Acceleration signal and then using modified double integration to get position signal. I also want to use modified control algorithm such as LQG optimal control or some robust controller instead of PID. Please let me know how can the FPGA module be configurable for performing such tasks. Better provide me some documents to start with.

Thanks again

Rahul

[Jochen]

Rahul,

do you want to use the  acceleration values to calculate the position displacement of the base where your motion stage is mounted? In general this should work but you will need to take into account that noise and other sources of inaccuracy could result in some inaccurracies in your position calculations.

With an R-Series board you can define any control algorithm that you like but you will have to implement it in integer/fixed point arithmetic as there is no floating point unit on the FPGA. I'm not aware of a ready to run example for an LQG algorithm but with Google I have found lot's of scientific abstracts about LQG and fixed point math so you will probably not have to reinvent the wheel.

Best regards,

Jochen