Embedded Online Optimization for Model Predictive Control at Megahertz Rates

Abstract

Faster, cheaper, and more power efficient optimization solvers than those currently offered by general-purpose solutions are required for extending the use of model predictive control (MPC) to resource-constrained embedded platforms. We propose several custom computational architectures for different first-order optimization methods that can handle linear-quadratic MPC problems with input, input-rate, and soft state constraints. We provide analysis ensuring the reliable operation of the resulting controller under reduced precision fixed-point arithmetic. Implementation of the proposed architectures in FPGAs shows that satisfactory control performance at a sample rate beyond 1 MHz is achievable even on low-end devices, opening up new possibilities for the application of MPC on embedded systems.

References in corpus (2)

• Convergence Rates of Inexact Proximal-Gradient Methods for Convex Optimization
• An ADMM Algorithm for Solving l_1 Regularized MPC

Cited by in corpus (34)

• OSQP: An Operator Splitting Solver for Quadratic Programs
• Efficient representation and approximation of model predictive control laws via deep learning
• High-Speed Finite Control Set Model Predictive Control for Power Electronics
• Prediction-Correction Algorithms for Time-Varying Constrained Optimization
• BLASFEO: basic linear algebra subroutines for embedded optimization
• Time-Varying Convex Optimization via Time-Varying Averaged Operators
• Energy-Efficient Real-Time Scheduling for Two-Type Heterogeneous Multiprocessors
• A General System for Heuristic Solution of Convex Problems over Nonconvex Sets
• Harmonic based model predictive control for set-point tracking
• The BLAS API of BLASFEO: optimizing performance for small matrices
• Nonlinear Predictive Control on a Heterogeneous Computing Platform
• Automatic Software and Computing Hardware Co-design for Predictive Control
• Deployment Architectures for Cyber-Physical Control Systems
• Improving Fast Dual Ascent for MPC - Part II: The Embedded Case
• Low-complexity method for hybrid MPC with local guarantees
• Learning Convex Optimization Control Policies
• Horizon-independent Preconditioner Design for Linear Predictive Control
• Prediction-Correction Splittings for Nonsmooth Time-Varying Optimization
• A Simple Effective Heuristic for Embedded Mixed-Integer Quadratic Programming
• Feedback and Time are Essential for the Optimal Control of Computing Systems
• Model predictive control of resistive wall mode for ITER
• A Penalty Method Based Approach for Autonomous Navigation using Nonlinear Model Predictive Control
• Implementation of model predictive control for tracking in embedded systems using a sparse extended ADMM algorithm
• Synthesis to Deployment: Cyber-Physical Control Architectures
• Exact and Heuristic Methods with Warm-start for Embedded Mixed-Integer Quadratic Programming Based on Accelerated Dual Gradient Projection
• Finite-word-length FPGA implementation of model predictive control for ITER resistive wall mode control
• ADMM for Block Circulant Model Predictive Control
• Prediction-Correction for Nonsmooth Time-Varying Optimization via Forward-Backward Envelopes
• Augmented Lagrangian Optimization under Fixed-Point Arithmetic
• Proportional-Integral Projected Gradient Method for Model Predictive Control
• Time-Varying Optimization: Algorithms and Engineering Applications
• Artificial-reference tracking MPC with probabilistically validated performance on industrial embedded systems
• Verification and Validation of Convex Optimization Algorithms for Model Predictive Control
• GoPRONTO: a Feedback-based Framework for Nonlinear Optimal Control