Polyhedral Control Design: Theory and Methods

Boris Houska,Matthias A. Müller,Mario E. Villanueva

2024-12-18

Abstract:In this article, we survey the primary research on polyhedral computing methods for constrained linear control systems. Our focus is on the modeling power of convex optimization, featured to design set-based robust and optimal controllers. In detail, we review the state-of-the-art techniques for computing geometric structures such as robust control invariant polytopes. Moreover, we survey recent methods for constructing control Lyapunov functions with polyhedral epigraphs as well as the extensive literature on robust model predictive control. The article concludes with a discussion of both the complexity and potential of polyhedral computing methods that rely on large-scale convex optimization.

Optimization and Control,Systems and Control

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: in modern robust control theories and methods, how to effectively represent and optimize sets that may be high - dimensional, especially in learning in uncertain environments and robust feedback control synthesis. Specifically, the paper explores the potential of polyhedral and polytopic computational tools in representing and optimizing high - dimensional convex sets. These tools can meet the modern set - based control requirements, especially in data - driven optimal control where there is significant uncertainty.

### Main Problem Summary:

1. **Representation and Optimization of High - Dimensional Sets**:
- Modern robust control theory relies on set - based computational tools for calculating safety margins in uncertain environments and robust feedback control synthesis.
- The key challenge is to develop tools that can effectively represent and optimize potentially high - dimensional sets.

2. **Application of Polyhedral and Polytopic Computational Tools**:
- The paper believes that polyhedral and polytopic computational tools are one of the most promising candidates for representing and optimizing high - dimensional convex sets.
- These tools utilize linear programming and related convex optimization techniques, which have been proven to be highly versatile and scalable.

3. **Specific Application Areas**:
- **Geometric Structure Computation**: Such as the computation of robust control invariant polytopes.
- **Control Lyapunov Functions (CLFs)**: Construct CLFs with polyhedral epigraphs.
- **Robust Model Predictive Control (MPC)**: There is extensive literature discussing convex - optimization - based robust MPC methods.

4. **Complexity and Potential**:
- The paper discusses the complexity and potential of polyhedral computational methods that rely on large - scale convex optimization, emphasizing the advantages of these methods in dealing with high - dimensional convex sets.

### Formula Representation:

- **Definition of Polyhedron**:
\[
P(F, y)=\{x\in\mathbb{R}^n\mid Fx\leq y\}
\]
where \(F\in\mathbb{R}^{f\times n}\) is a matrix and \(y\in\mathbb{R}^f\) is a vector.

- **Gale's Theorem**:
The polyhedron \(P(F, y)\) is non - empty if and only if there exists a vector \(z\in\mathbb{R}^f\) satisfying:
\[
z\geq0,\quad F^{\top}z = 0,\quad y^{\top}z<0
\]

- **Minkowski - Weyl Theorem**:
The polyhedron \(P(F, y)\) is non - empty if and only if there exist matrices \(Y\) and \(Z\) such that:
\[
P(F, y)=\text{conv}(Y)+\text{cone}(Z)
\]
where \(\text{conv}(Y)\) represents the convex hull of the column vectors of \(Y\) and \(\text{cone}(Z)\) represents the convex cone generated by the column vectors of \(Z\).

Through these methods and tools, the paper aims to provide readers with a comprehensive polyhedral method toolkit to meet a wide range of set - based control challenges.