Introduction To The Pontryagin Maximum Principle For Quantum Optimal Control Guide

$i\hbar \frac\partial\partial t |\psi(t)\rangle = H(t) |\psi(t)\rangle$

The PMP is a rapidly growing field in quantum optimal control, and there are several future directions:

The is a fundamental method in Optimal Control Theory (OCT) used to find the best way to manipulate quantum systems, such as preparing specific quantum states or minimizing energy loss . It provides the necessary conditions to determine "open-loop" control laws—instructions for external fields like lasers—without needing real-time feedback. Core Concepts of PMP For a quadratic penalty ( \mathcalL(u) = \frac\lambda2

The PMP dictates that the optimal control ( u_k^*(t) ) maximizes ( \mathcalH_P ) pointwise in time. For a quadratic penalty ( \mathcalL(u) = \frac\lambda2 u_k^2 ), this gives:

PMP says the optimal control switches between extreme values (bang-bang) unless singular. This article introduces the PMP from first principles,

Quantum Optimal Control (QOC) is the science of steering quantum systems from an initial state to a target state using external electromagnetic fields. While gradient-based algorithms are common, they often act as "black boxes." The Pontryagin Maximum Principle (PMP) provides the fundamental mathematical backbone for understanding why optimal control laws exist and what their structure must be. This article introduces the PMP from first principles, translates its core concepts into the language of Hilbert spaces and unitary evolution, and demonstrates how it leads to the powerful framework of and Gradient Ascent control strategies.

Where ( O ) is a target observable (e.g., projector onto a desired state) and ( \mathcalL ) penalizes large or oscillatory controls. at each instant

The Pontryagin Maximum Principle answers a critical question:

Quantum optimal control seeks to manipulate a quantum system via external fields—such as lasers, radiofrequency pulses, or microwave signals—to achieve a desired target behavior while minimizing a specific cost functional.

In short: Optimal controls are those that, at each instant, greedily maximize the instantaneous Hamiltonian of the costate.