[R] curve fitting question

[tom wright]

I'd appreciate some direction here.
I have a model for a system with two independant variables (i1,i2) and
one dependant variable (d).
I have experimental data recorded at multiple levels of the dependant
variable (x).
I need to work out the values for the independant variables that best
fit the experimental data recorded for all the dependant variables.
I assume I'm going to need the glm() function but I'd really appreciate
some pointers in how to actually do this.
Do I need to calculate the fits for a range of values of i1 and i2 for
each value of d and then look to see what values of i1 and i2 give the
best fit over all the experimental data or is there a way of doing this
automatically?
Please feel free to point me at other reading sources.

Many thanks
Tom