25 April 2019 in IEEE Control Systems Letters
Model Predictive Control (MPC) is an enabling technology in applications requiring controlling physical processes in an optimized way under constraints on inputs and outputs. However, in MPC closed-loop performance is pushed to the limits only if the plant under control is accurately modeled; otherwise, robust architectures need to be employed, at the price of reduced performance due to worst-case conservative assumptions. In this paper, instead of adapting the controller to handle uncertainty, we adapt the learning procedure so that the prediction model is selected to provide the best closed-loop performance. More specifically, we apply for the first time the above “identification for control” rationale to hierarchical MPC using data-driven methods and Bayesian optimization.
14 January 2021 in International Journal of Adaptive Control and Signal Processing
21 November 2023 in IEEE Control Systems Letters
Is it possible to understand the intricacies of a dynamical system not solely from its input/output pattern, but also by observing the behavior of other systems within the same class? This central question drives the study presented in this paper. In response to this query, we introduce a novel paradigm for system identification, addressing two primary tasks: one-step-ahead prediction and multi-step simulation. Unlike conventional methods, we do not directly estimate a model for the specific system. Instead, we learn a meta model that represents a class of dynamical systems. This meta model is trained on a potentially infinite stream of synthetic data, generated by simulators whose settings are randomly extracted from a probability distribution. When provided with a context from a new system–specifically, an input/output sequence–the meta model implicitly discerns its dynamics, enabling predictions of its behavior. The proposed approach harnesses the power of Transformers, renowned for their in-context learning capabilities. For one-step prediction, a GPT-like decoder-only architecture is utilized, whereas the simulation problem employs an encoder-decoder structure. Initial experimental results affirmatively answer our foundational question, opening doors to fresh research avenues in system identification.
15 May 2020 in IEEE Control Systems Letters
This paper introduces a novel neural network architecture, called Integrated Neural Network (INN), for direct identification of nonlinear continuous-time dynamical models in state-space representation. The proposed INN is used to approximate the continuous-time state map, and it consists of a feed-forward network followed by an integral block. The unknown parameters are estimated by minimizing a properly constructed dual-objective criterion. The effectiveness of the proposed methodology is assessed against the Cascaded Tanks System benchmark.
1 January 2020 in IFAC-PapersOnLine
26 October 2021 in IEEE Transactions on Automatic Control
Identification of dynamical systems from low-resolution quantized observations presents several challenges because of the limited amount of information available in the data and since proper algorithms have to be designed to handle the error due to quantization. In this article, we consider identification of infinite impulse response models from quantized outputs. Algorithms both for maximum-likelihood estimation and Bayesian inference are developed. Finally, a particle-filter approach is presented for recursive reconstruction of the latent nonquantized outputs from past quantized observations.