Author: Martin Enqvist

Abstract: Modern ships are equipped with various control and automation systems and the tuning of these systems usually requires accurate mathematical models, for example, second-order modulus models. System identification is a natural approach in this application, but a main challenge is that the time for data collection is often quite limited during the commissioning of a new ship, making experiment design a key tool. Here, a dictionary-based approach to experiment design for ships will be discussed. This approach presents a systematic way of choosing the most informative combination of independent experiments out of a predefined set of candidates. This idea is quite general but is here tailored to an instrumental variable (IV) estimator with zero-mean instruments, which is an estimator well suited to deal with parameter estimation for second-order modulus models. The method is evaluated using both simulated and real data, the latter from a small model ship as well as from a full-scale vessel. Further, a standard motion-planning problem is modified to account for the prior-made choice of information-optimal sub-experiments, which makes it possible to obtain a plan for the complete experiment in the form of a feasible trajectory.
Slides – Paper

Authors: Stéphane Victor, Mohamed Hajjem, Patrick Lanusse, Pierre Melchior, Lara Thomas

Abstract: This paper proposes a method to design a real-time wind turbulence simulator. The objective of such a simulator is to make dynamical models more accurate in order to develop more robust controls, especially in the case of mechanical systems operating outdoors such as tracking antennas. The models used to generate a random wind speed are based on the wind spectral characteristics rather than time domain ones. Indeed, turbulence is a stochastic and non-stationary process corresponding to the short-term component of wind and therefore is difficult to model in time domain. Wind spectral characteristics are described by the power spectral density whose approximation is used in the real-time simulator to reproduce wind behavior. The von Karman model is the most commonly used model to approximate the power spectral density of wind turbulence. However, originally designed for aircraft, and therefore for high altitudes and moving systems, tracking antennas are used under different conditions: low altitude and slow moving systems. This makes the von Karman model less precise in middle and high range frequency. Consequently, there is a need in more accurately modeling wind turbulence under these specific conditions. Other models such as Cole-Cole and Davidson-Cole fractional models are compared. and an improved fractional Cole-Cole model is proposed which has only four parameters. Finally, an accurate turbulence wind speed generator is proposed.
Slides – Paper

Authors: Hugo Koide, Guillaume Mercère, Jérémy Vayssettes

Abstract: Recursive parameter estimation methods tailored to errors-in-variables (EIV) systems have long been focused on extending the classical total least squares (TLS) method based on the singular value decomposition (SVD). Such methods find ways to recursively update the singular vectors of the augmented data matrix of noise-ridden inputs and outputs in a linear EIV system. In parallel to the fields of system identification, signal processing, and automatic control, however, an iterative TLS solution grounded in the nonlinear optimization of the EIV equation has been popularized in the field of geodetics. This opens doors to the development of a new recursive TLS formulation, which closely resembles the classical recursive least squares (RLS) formulation and provides real-time knowledge of the estimated input and output measurement biases. Experiments show that RTLS methods based on said nonlinear Lagrange optimization have more robust properties when compared to SVD-based solutions when catering to ill-posed problems and data with non-uniform noise characteristics.
Slides – Paper

Authors: Pauline Kergus, Ion Victor Gosea, Mihaly Petreczky

Abstract: This work brings together the moment matching approach based on Loewner functions and the classical Loewner framework based on the Loewner pencil in the case of bilinear systems. Loewner functions are defined based on the bilinear Loewner framework, and a Loewner equivalent model is produced using these functions. This model is composed of infinite series that needs to be truncated in order to be implemented in practice. It is shown that the moment matching procedure based on the proposed Loewner functions and the classical interpolatory bilinear Loewner framework both result in approximate Loewner equivalent models, the main difference being that the latter preserves bilinearity at the expense of a higher order.
Slides – Paper

Authors: Fatima Zahra Boutourda, Régis Ouvrard, Thierry Poinot, Driss Mehdi, Fouad Mesquine, Vincent Jauzein, Eloise De Tredern

Abstract: In wastewater treatment, understanding and modeling the nitrification process is crucial for effective control implementation. However, the complexity of this process makes it challenging to create simplified models. This study introduces an innovative method for estimating linear parameter varying (LPV) models in the context of biological nitrification processes. The research focuses on the development of a continuous-time LPV model that replicates the behavior observed in the nitrification system. The methodology adopts the reinitialized partial moment approach within a global identification framework. The resultant LPV model is structured to capture the dynamics of the biological nitrification process, considering various factors like flow rates, feed concentrations and environmental regulations. Application of this approach to measured data from a wastewater treatment plant, demonstrates its effectiveness in accurately estimating the LPV model parameters. The results not only offer valuable insights into the dynamics and the nonlinear behaviour of the nitrification process but also contribute to the design and optimization of wastewater treatment plants, particularly those employing submerged aerated nitrifying biofilters.
Slide – Paper

Authors: Yan Monier, Bruno Denis, Gregory Faraut, Nabil Anwer

Abstract:  La modélisation de systèmes cyber-physiques est un des challenges principaux de l’industrie 4.0. Nous nous intéressons dans ces travaux à la modélisation de systèmes cyber-physiques par identification et plus particulièrement à l’identification de systèmes hybrides réactifs déterministes sous la forme d’un automate hybride. Dans la littérature, le problème d’identification de système à un automate hybride peut être divisé en 5 sous problèmes résolus de manière séquentielle. Nous nous intéressons dans ces travaux à proposer une évolution de ce framework afin de lever certaines limites empêchant une optimisation du modèle identifié dans son ensemble.

Slides – Paper

Frequency Domain

Authors: Omar Arahbi, Benoît Huard, Jean-Denis Gabano, Thierry Poinot

Abstract:  Electrochemical Impedance Spectroscopy (EIS) is a useful tool for selecting a pertinent Equivalent Circuit Model (ECM) of a Li-ion battery. Impedance model is designed to describe low, middle and high frequency electrochemical processes involved. When considering low frequency restricted in the Warburg zone, diffusion impedance is modeled thanks to a Constant Phase Element (CPE) which behaves as a fractional integrator of order n close to 0.5. Phenomena observed in middle frequency are described using specific circuits called Zarc which consist in connecting a CPE in parallel with a resistor. Therefore, the global impedance model is characterized by non integer order operators and parameters can be estimated by a Complex Non linear Least Squares (CNLS) algorithm which requires a proper initialization in order to guarantee the convergence to a global optimum. The paper presents a method to analyze EIS data measurements in order to select automatically the number of middle frequency Zarc circuits required (one or two) and to initialize properly the CNLS algorithm. The method is validated using experimental open source EIS data.
Slide – Paper

Authors: Alain Uwadukunze, Xavier Bombois, Marco Forgione, Marion Gilson, Marie Albisser

Abstract:  Designing controllers for real life processes can pose significant challenges due their non-linear nature. The goal of this talk is to present a model-based feedforward controller design for nonlinear systems in an open-loop framework. We propose the use of block-oriented models to identify the model of the true system then design a feed-forward controller based on the identified model. An iterative identification and controller design procedure is also proposed in a case where the initially identified model does not allow to design a suitable controller. The procedure is applied on the design of a feed-forward controller for a Wiener system. 
Slides – Paper

Authors: Lucas Gruss, Mohamed Yagoubi, Maxime Thieffry, Philippe Chevrel

Abstract:  L’estimation de l’état étendu d’un modèle de réacteur nucléaire est traitée par l’implémentation d’un estimateur à horizon mouvant. L’état ainsi que des paramètres d’intérêt du modèle sont identifiés simultanément. Les spécificités du modèle (raideur, non-linéaire) nécessitent une implémentation minutieuse et des méthodes de résolution adaptées. La transcription du problème de commande optimale est effectuée par collocation directe. L’estimateur ainsi synthétisé est comparé à un filtre de Kalman étendu, qui ne produit pas d’estimées satisfaisantes des états et des paramètres.
Slides – Paper