08-04-2009 09:43 AM
I am having a little trouble getting started. I am going to build a pretty basic engine control system. I need to be able to move a linear actuator (~4" stroke) to control the throttle and read RPM (in frequency) as the feedback.
The system should have the ability to command the position of the throttle or command an RPM. I had originally pictured a variable voltage output with soft PID, but everything I see is for servo/stepper controls? I think a windows based PID will be fine, these engines don't accelerate quickly and the RPM control does not need precision like a RT system would deliver.
I am only familiar with Labview 8.6 and PXI/SCXI systems running windows. I would not be opposed to going to a CRIo type of system, but I don't think it is necessary, and the system needs to be running in a fairly short amount of time.
I don't have an actuator picked out, but am open to suggestions. I'm not sure of the force requirements yet, guessing between 5 and 50 lbs force, but it could be less.
08-05-2009 06:28 AM - edited 08-05-2009 06:31 AM
1. I need to control the speed of a jet engine.
2. The actuator will connect to the valve on the FPMU. This is basically like connecting to the throttle blade of a mechanically injected piston engine (Or at least thats what they tell me, I'm a piston guy myself). A stepper/servo motor with a bell crank would also work, but I think an actuator would be slightly more simple.
3. I would say neither lab or real world. The initial development is for an engine mounted outdoors on a stand used to blow air over a test article. This thing will evolve, perhaps to control on air speed, as well as to go on a different engine that will be tested. But for now, moving the throttle is the absolute necessity and controlling on RPM is the secondary goal. The rest of the stuff can come later, and will probably be started from scratch with lessons learned from this first phase.
We will also, in this and all follow on phases, monitor EGT, oil pressure, fuel pressure, etc.. through and SCXI 1102C or similar, and control on off relays with a PXI6713 or similar based on either user inputs or value monitoring. That part I'm comfortable with, its just moving the throttle blade that I'm a little stuck on.
08-05-2009 12:25 PM
As far as the actuator goes, I have used the Parker ET series. They have a model for pretty small loads, but it still may be overkill for your application. You can use their servo motor, drive, and control and talk to Labview from their ActiveX library, or put your own stepper or servo on it. The Parker Aries controller is a good drive/controller for small servos.
08-05-2009 12:25 PM
These electric cylinders may be overkill for your application. I have used them a couple times, these things are nearly indestructable. You can order them with an encoder and servo motor for closed loop positioning control.
08-05-2009 02:59 PM
08-05-2009 06:38 PM
I think the PID examples use electric motors commonly because they're easy to adequately model and understand. If you can model your system, you can make use of the Control Design and Simulation Toolkit (and maybe the System Identification Toolkit to get the model in the first place) to do your control.
08-05-2009 08:47 PM - edited 08-05-2009 08:48 PM
08-06-2009 09:28 AM
What kind of hardware would I need to use the toolkits you referenced? I know a little bit about those two, and I think this control will be simple enough to do something like that, but I'm not sure of the NI hardware that would go with that?
That is the vi I started with also, and no matter what I do it pegs the processor usage on my little computers. I guess I just need bigger computers. As far as understanding the ACR, I would say I understand it enough to be dangerous.