Augmented Lagrange method for optimal control problems of parabolic equation with state constraints

Weilong You,Fu Zhang

2024-11-28

Abstract:The augmented Lagrange method is employed to address the optimal control problem involving pointwise state constraints in parabolic equations. The strong convergence of the primal variables and the weak convergence of the dual variables are rigorously established. The sub-problems arising in the algorithm are solved using the Method of Successive Approximations (MSA), derived from Pontryagin's principle. Numerical experiments are provided to validate the convergence of the proposed algorithm.

Optimization and Control,Numerical Analysis

What problem does this paper attempt to address?

This paper attempts to solve the optimal control problem of parabolic equations with state constraints. Specifically, the article focuses on how to handle pointwise state constraints in parabolic equations. The main challenge in the numerical solution of such problems is that the Lagrange multiplier belongs to the Radon measure space \(M(\Omega_T)\), which makes the Lagrange multiplier less regular and thus brings difficulties to numerical methods.

To solve this problem, the author proposes to use the Augmented Lagrange Method. This method solves the problem by adding the inequality state constraint \(y \leq \psi\) as a penalty term to the objective function. This can avoid directly dealing with the Lagrange multiplier with low regularity and can solve the optimal control problem more effectively.

The main contributions of the article include:
1. **Strict convergence analysis**: Prove the strong convergence of the original variables and the weak convergence of the dual variables.
2. **Sub - problem solving method**: Use the Method of Successive Approximations (MSA) based on Pontryagin's Principle to solve the sub - problems in the Augmented Lagrange Method.
3. **Numerical experiment verification**: Provide numerical experiment results to verify the effectiveness and convergence of the proposed algorithm.

### Formula summary
- Objective function:
\[
J(y, u, v)=\frac{1}{2}\|y(\cdot, T)-y_d\|_{L^2(\Omega)}^2+\frac{\alpha}{2}\|u\|_{L^2(\Omega_T)}^2+\frac{\beta}{2}\|v\|_{L^2(\Sigma_T)}^2
\]
- Augmented Lagrange function:
\[
L_{\rho}(y, u, v, \mu)=J(y, u, v)+\frac{\rho}{2}\|y - \psi+\frac{\mu}{\rho}-P_K(y - \psi+\frac{\mu}{\rho})\|_{L^2(\Omega_T)}^2-\frac{1}{2\rho}\|\mu\|_{L^2(\Omega_T)}^2
\]
where \(P_K\) is the projection operator from \(L^2(\Omega_T)\) to \(K = \{y \in L^2(\Omega_T)\mid y \leq 0\}\).

Through these methods and formulas, the author successfully solves the optimal control problem of parabolic equations with state constraints and verifies its effectiveness through numerical experiments.