The simplest way is to turn it to an unconstrained optimization problem that is T
t= 0
βtU(f(kt) kt+ 1 )!max