On Fri, 19 Oct 2018 at 19:14, Nicklas Karlsson
<[email protected]> wrote:

> I found state space feedback and bode, nyquist diagram are useful also for 
> state feedback but then state space model is available

I have never studied any of that stuff. Which is not to dismiss it, I
did Physics at college and am a fan of analysis. But in my day job I
basically program controllers (engine management systems) and the
theory isn't much help. Partly because exploring the configuration
space is impossible in finite time, and partly because exploring the
response time gets very expensive, very fast.

I work mainly on the idle speed controller. It's PID. Except that
there are separate gains for every gear, and for various clutch
engagement states. And every P, I and D is an 8x8 map which varies by
idle speed error and engine temperature.

Many of the points in those curves either are not stable enough to
test (400 rpm below setpoint) or don't stay still (-30C coolant temp)
so we have to throw theory out the window (whilst acknowledging what
it has to tell us) and just turn the knobs in real time listening for
the result. (like a trombonist or violinist)

-- 
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916


_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to