ABSTRACT
Central difference: e(k + 1)- e(k-1) 2T Although the central difference method provides the best results since it "averages" over two sample periods, a future error value is needed and so from a programming perspective it is not that useful. The same is true for the forward difference approximation. In general, then, the backward difference is commonly used to approximate the derivative term in our PID algorithm.
