Hierarchical Causal Abduction: A Foundation Framework for Explainable Model Predictive Control
For operators of safety-critical infrastructure using nonlinear MPC, HCA provides faithful explanations that improve trust and enable deployment, addressing a key bottleneck in opaque control systems.
The paper introduces Hierarchical Causal Abduction (HCA), a framework that combines physics-informed reasoning, optimization evidence, and temporal causal discovery to generate interpretable explanations for nonlinear MPC actions. HCA achieves 53% higher explanation accuracy than LIME (0.478 vs. 0.311) across three control domains without per-domain tuning, and up to 0.88 with calibration.
Model Predictive Control (MPC) is widely used to operate safety-critical infrastructure by predicting future trajectories and optimizing control actions. However, nonlinear dynamics, hard safety constraints, and numerical optimization often render individual control moves opaque to human operators, undermining trust and hindering deployment. This paper presents Hierarchical Causal Abduction (HCA), which combines (i) physics-informed reasoning via domain knowledge graphs, (ii) optimization evidence from Karush--Kuhn--Tucker (KKT) multipliers, and (iii) temporal causal discovery via the PCMCI algorithm to generate faithful, human-interpretable explanations for control actions computed by nonlinear MPC. Across three diverse control applications (greenhouse climate, building HVAC, chemical process engineering) with expert validation, HCA improves explanation accuracy by 53\% over LIME (0.478 vs. 0.311) using a single set of cross-domain parameters without per-domain tuning; domain-specific KKT-threshold calibration over 2--3 days further increases accuracy to 0.88. Ablation studies confirm that each evidence source is essential, with 32--37\% accuracy degradation when any component is removed, and HCA's ranking-and-validation methodology generalizes beyond MPC to other prediction-based decision systems, including learning-based control and trajectory planning.