Nonautonomous saddle node bifurcations commonly appear in mathematical models; however, an optimal control theory tailored to such systems is not yet well established. In this talk, we present conceptual work on the nonautonomous saddle node normal form. We develop an approach that reformulates the

optimal control problem into an unconstrained minimization problem by exploiting a connection between bifurcation and rate–induced overshoots. The key component is a rate function that assigns the critical rate to every nonautonomous forcing. We show that in the saddle node normal form, the critical rate is

unique, and hence the rate function is well defined.

While this reformulation is valid in any case, the rate function is often unknown. To resolve this, we additionally present an approximation of the rate function that leads to an analytic expression of the approximated control. This makes the approach appealing in situations in which we need repeated solves of the optimal control problem such as in cases that induce a sampling-based approach.

Finally, we show that the reformulated problem can be used to recover a generalization of the Euler-Lagrange equations.