Optimizing the Energetics of the Finite-time Driving of Field Theories

Abstract

The phase transitions for many-body systems have been understood using field theories. A few canonical physical model classes encapsulate the underlying physical properties of a large number of systems. The finite-time driving of such systems and associated optimal energetic costs have not been investigated yet. We consider two universality classes Model A and Model B, that describe the dynamics for the non-conserved and conserved scalar order parameters respectively. Here, using the recent developments in stochastic thermodynamics and optimal transport theory, we analytically compute the optimal driving protocols by minimizing the mean stochastic work required for finite-time driving. Further, we numerically optimize the mean and variance of the stochastic work simultaneously. Such a multi-objective optimization is called a Pareto optimization problem and its optimal solution is a Pareto front. We discover a first-order Pareto phase transition in the Pareto front. Physically, it corresponds to the coexistence of two classes of optimal driving protocols analogous to the liquid-gas coexistence for the equilibrium phase transition. Our framework sheds light on the finite-time optimal driving of the fields and the trade-off between the mean and fluctuations of the optimal work.