Analysis and design of model predictive control frameworks for dynamic operation -- An overview

Johannes Köhler,Matthas A. Müller,Frank Allgöwer

DOI: https://doi.org/10.1016/j.arcontrol.2023.100929

2024-01-09

Abstract:This article provides an overview of model predictive control (MPC) frameworks for dynamic operation of nonlinear constrained systems. Dynamic operation is often an integral part of the control objective, ranging from tracking of reference signals to the general economic operation of a plant under online changing time-varying operating conditions. We focus on the particular challenges that arise when dealing with such more general control goals and present methods that have emerged in the literature to address these issues. The goal of this article is to present an overview of the state-of-the-art techniques, providing a diverse toolkit to apply and further develop MPC formulations that can handle the challenges intrinsic to dynamic operation. We also critically assess the applicability of the different research directions, discussing limitations and opportunities for further research.

Systems and Control,Optimization and Control

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve

The paper aims to address the design and analysis issues of the Model Predictive Control (MPC) framework in dynamic operations. Specifically, the paper focuses on the following three challenges:

1. **Non-steady-state operation** (C.1): The system is not expected to reach a steady state but needs to track time-varying reference trajectories or operate within some general sets.
2. **Unpredictable online changes** (C.2): The desired operation mode changes unpredictably online.
3. **Inability to directly specify reference trajectories** (C.3): The desired operation mode cannot be directly specified by given state setpoints or trajectories.

### Background of Dynamic Operations

Dynamic operations are becoming increasingly important in many emerging control applications, such as motion planning or trajectory optimization problems in robotics, aerospace, autonomous driving, and HVAC systems in buildings. In these applications, the control objective is often not a simple steady-state stabilization problem but involves the optimization of economic performance indicators (e.g., energy consumption).

### Main Contributions of the Paper

The paper provides an overview of recent advances in the design and analysis of MPC frameworks capable of addressing the challenges of dynamic operations. Specific contributions include:

1. **Classical system theory properties**: MPC design tools and methods for general nonlinear systems to ensure recursive feasibility, constraint satisfaction, and stability/performance.
2. **Artificial reference trajectories**: How to handle infeasible and online-changing reference trajectories.
3. **Economic performance optimization**: How to directly optimize economic performance indicators.
4. **Simplified MPC frameworks**: Analysis methods for MPC frameworks that do not require stabilizing terminal components.

### Structure of the Paper

1. **Introduction**: Introduces the basic principles of MPC and its application background in dynamic operations.
2. **Preliminaries**: Explains the basic concepts of stable MPC design, including recursive feasibility and stability.
3. **Terminal component design**: Designs terminal costs and terminal sets for non-steady-state tracking problems.
4. **Artificial reference trajectories**: Methods for handling infeasible and online-changing reference trajectories.
5. **Economic MPC**: Methods for directly optimizing economic performance indicators.
6. **Simplified MPC frameworks**: Analysis methods for MPC frameworks that do not require stabilizing terminal components.
7. **Discussion**: Summarizes the advantages, limitations, and future research directions of different methods.

### Key Challenges

- **Non-steady-state operation** (C.1): The system is not expected to reach a steady state but needs to track time-varying reference trajectories.
- **Unpredictable online changes** (C.2): The desired operation mode changes unpredictably online.
- **Inability to directly specify reference trajectories** (C.3): The desired operation mode cannot be directly specified by given state setpoints or trajectories.

### Solutions

- **Terminal component design**: Ensures local stability by designing terminal costs and terminal sets through linearization and LQR methods.
- **Artificial reference trajectories**: Introduces artificial reference trajectories to handle infeasible and online-changing reference trajectories.
- **Economic MPC**: Directly optimizes economic performance indicators without relying on traditional stabilizing terminal costs.
- **Simplified MPC frameworks**: Studies MPC frameworks that do not require stabilizing terminal components, simplifying the design process.

Through these methods, the paper provides a comprehensive set of tools and methods for MPC design in dynamic operations, addressing the limitations of traditional MPC in handling complex dynamic operations.