efi control algorithm

Mike Klopfer klopfer at eagle.natinst.com
Fri May 19 01:55:57 GMT 1995


An excerpt from a previous post I believe by pecampbe at mtu.edu:
>2. "The" model recommended for electronic fuel injection seems to be the
>mean value engine model. I have two questions related to it:
>
>a. How do you measure the parameters required for the model? The "perfect"
>solution of course would be some simple user-entry parameters (number of
>cylinders, displacement size, overbore size, CFI/SEFI/EFI injection type),
>and then let the engine computer determine the rest online.
>
>b. How would you convert it from a model to a control algorithm? I understand
>that there are 3 inputs and 2 states..but the next step is to derive some
>physical parameters that you wish to optimize based on the model. It does
>not appear that anyone has ground out the mathematics yet except Mike
>Klopfer (Jan 7, 1995) who seems to have made a pretty good stab with a similar
>algorithm.
>

The control equations are in the Jan post. There are two parts to it, the 
first part is the equation:

on(k)= (injSt(k)-t0) phiDes - 
       (f_b(k)/f_a(k)) {efp(k)-[(1-f_a(k))/f_b(k)](injSt(k)-t0) phiDes}

     = [(injSt(k)-t0) phiDes/f_a(k)]-[(f_b(k)efp(k))/f_a(k)]

The second part involves determining efp(k) using the Kalman equations.

The following is some theorizing about how one might calibrate the parameters
needed. The most important parameter seems to me to be the map of the steady 
state injector on time injSt. This is a table of the injector pulse width at
various throttle angles and rpms that is needed provide a stoichiometric ratio
of fuel to air.


More information about the Diy_efi mailing list