In this paper a novel feeding strategy is suggested for bioreactor pro
cesses. The strategy, referred to as the modified linear feeding strat
egy (MLFS), is shown to overcome the disadvantages of the feedback exp
onential feeding strategy and the linear feeding strategy, resulting i
n a fast response in substrate concentration for a large range of valu
es of initial conditions. It is also shown that one of the main featur
es of the MLFS is that it effectively decomposes the closed-loop syste
m into a fast and a slow subsystem. This property is used to prove tha
t the control input cannot saturate in the case when an adaptive versi
on of MLFS is used to control the processes with Haldane kinetics. In
the latter case, the MLFS is also shown to effectively help in assurin
g the stability of the closed-loop system. Applications of the MLFS in
both nonadaptive and adaptive cases are illustrated through computer
simulations. (C) 1997 Elsevier Science Ltd.