Integrating Bayesian methods with neural network--based model predictive control: a review
This review identifies gaps in evaluating Bayesian techniques for MPC, calling for standardized benchmarks to improve reliability in control systems, but it is incremental as it synthesizes existing studies without new results.
The review assesses the integration of Bayesian methods with neural network-based model predictive control (MPC) to address uncertainty, finding that reported performance and robustness gains are fragmented due to inconsistent baselines and limited reliability analyses.
In this review, we assess the use of Bayesian methods in model predictive control (MPC), focusing on neural-network-based modeling, control design, and uncertainty quantification. We systematically analyze individual studies and how they are implemented in practice. While Bayesian approaches are increasingly adopted to capture and propagate uncertainty in MPC, reported gains in performance and robustness remain fragmented, with inconsistent baselines and limited reliability analyses. We therefore argue for standardized benchmarks, ablation studies, and transparent reporting to rigorously determine the effectiveness of Bayesian techniques for MPC.