ReLiNet: Stable and Explainable Multistep Prediction with Recurrent Linear Parameter Varying Networks

Alexandra Baier, Decky Aspandi, Steffen Staab

Introduction

Predicting the future states of dynamical systems—such as maritime vessels or industrial robotic arms—is a fundamental challenge in automation and control. While deep learning models like Long Short-Term Memory (LSTM) networks have set new benchmarks for accuracy, they are often criticized as "black boxes" that lack formal stability guarantees and interpretability. Conversely, traditional control-theoretic models like switched linear systems offer mathematical elegance and stability but frequently fail to capture the complex nonlinearities of real-world data.

This paper introduces ReLiNet (Recurrent Linear Parameter-Varying Network), a novel architecture designed to bridge this gap. ReLiNet provides high-accuracy multistep predictions that are formally stable and inherently explainable by design.

Methodology

ReLiNet employs a "master-slave" architecture to combine the representational power of recurrent neural networks (RNNs) with the transparency of linear models.

The Master-Slave Architecture

The "Master" component is an RNN that processes control inputs to generate time-varying parameters (matrices) for a "Slave" linear model. The Slave model then uses these parameters to predict the next state based on the current state and control inputs. Because the final prediction is the result of a linear operation, the model is explainable by default; the contribution of each input to the output can be derived mathematically without relying on post-hoc approximation tools like LIME or SHAP.

Formal Stability with StableReLiNet

To address the risk of error accumulation in multistep predictions, the authors propose StableReLiNet. By enforcing specific mathematical constraints—ensuring system matrices are simultaneously diagonalizable and bounding their eigenvalues via a $tanh$ activation—the model is proven to be globally exponentially stable. This ensures that the model's outputs remain bounded and converge toward a steady state under constant inputs, a critical requirement for safety-critical applications.

Experimental Evaluation

The models were evaluated on two complex benchmarks: a Ship Maneuvering simulation (including an out-of-distribution set) and a real-world Industrial Robot joint angle dataset.

Accuracy and Generalization

ReLiNet achieves Normalized Root Mean Squared Error (NRMSE) scores comparable to state-of-the-art LSTMs. Notably, in the Ship OOD task, ReLiNet outperformed the LSTM across all prediction horizons, suggesting that its underlying linear structure provides superior generalization to unseen conditions compared to standard recurrent architectures.

NRMSE performance across varying prediction horizons (h) for the SHIP and ROBOT datasets.

Explainability and Faithfulness

A key advantage of ReLiNet is its "faithfulness." In explainable AI, infidelity measures how much an explanation deviates from the model's actual logic. ReLiNet achieves an infidelity score of zero, meaning its explanations are perfectly faithful representations of its internal mechanics. In contrast, post-hoc methods like LIME applied to LSTMs showed significantly higher infidelity, particularly on the ship datasets.

Explanation quality metrics (Infidelity, Lipschitz, and Similarity) comparing ReLiNet to LIME.

Discussion and Conclusion

The results demonstrate that ReLiNet successfully balances performance and safety. While StableReLiNet incurs a measurable "safety tax" in terms of accuracy compared to its unconstrained counterpart, it remains significantly more accurate than traditional linear baselines while providing rigorous stability guarantees. ReLiNet represents a significant advancement toward trustworthy AI in physical systems, proving that transparency and predictive power are not mutually exclusive.

Summary updated

Abstract

Multistep prediction models are essential for the simulation and model-predictive control of dynamical systems. Verifying the safety of such models is a multi-faceted problem requiring both system-theoretic guarantees as well as establishing trust with human users. In this work, we propose a novel approach, ReLiNet (Recurrent Linear Parameter Varying Network), to ensure safety for multistep prediction of dynamical systems. Our approach simplifies a recurrent neural network to a switched linear system that is constrained to guarantee exponential stability, which acts as a surrogate for safety from a system-theoretic perspective. Furthermore, ReLiNet's computation can be reduced to a single linear model for each time step, resulting in predictions that are explainable by definition, thereby establishing trust from a human-centric perspective. Our quantitative experiments show that ReLiNet achieves prediction accuracy comparable to that of state-of-the-art recurrent neural networks, while achieving more faithful and robust explanations compared to the model-agnostic explanation method of LIME.
BibTeX
@inproceedings{baier2023relinet,
  author = {Alexandra Baier and Decky Aspandi and Steffen Staab},
  title = {{ReLiNet: Stable and Explainable Multistep Prediction with Recurrent Linear Parameter Varying Networks}},
  booktitle = {International Joint Conference on Artificial Intelligence},
  year = {2023},
  doi = {10.24963/ijcai.2023/385},
}
Figure