Search

State estimation has a pivotal role in several applications, including but not limited to advanced control design. Especially when dealing with nonlinear systems state estimation is a nontrivial task, often entailing approximations and challenging fine-tuning phases. In this work, we propose to overcome these challenges by formulating an in-context state-estimation problem, enabling us to learn a state estimator for a class of (nonlinear) systems abstracting from particular instances of the state seen during training. To this end, we extend an in-context learning framework recently proposed for system identification, showing via a benchmark numerical example that this approach allows us to (i) use training data directly for the design of the state estimator, (ii) not requiring extensive fine-tuning procedures, while (iii) achieving superior performance compared to state-of-the-art benchmarks.

Increasing the manufacturing resource efficiency requires pushing the process limits by customizing the machining parameters to the job at hand. Such optimizations performed usually based on empirical methods and experience due to the sheer complexity of physical modeling. Current research presents a system comprised of advanced machine learning methods, where model-based optimization employed to simultaneously create a digital twin and optimize the process as a self-learning virtual operator using a case study with the EDM processes.

Providing the users information on the energy consumed in the household at the appliance level is of major importance for increasing their awareness of their consumption behavior. In this paper, we propose a technique based on Kalman filters to estimate the devices’ consumption patterns from aggregate readings, i.e., to solve the so called disaggregation problem. The method is suited for on-line disaggregation and the proposed results show that it is robust against modelling errors and unmodelled appliances.

A Least-Squares Support Vector Machine (LS-SVM) estimator, formulated in the frequency domain is proposed to identify linear time-varying dynamic systems. The LS-SVM aims at learning the structure of the time variation in a data driven way. The frequency domain is chosen for its superior robustness w.r.t. correlated errors for the calibration of the hyper parameters of the model.

Airborne 137Cs level in France is not decreasing significantly anymore (mean yearly value around 0,25 µ Bq.m-3contrarily to what was noticed in the past decades. This observation points out the role of processes that delay the atmospheric cleaning and participate to the persistence of radionuclide in the air at ground-level after a deposit, in the frame of medium as well as long-dated post-accidental contexts. The current background level also yields to consider 137Cs in the atmosphere as a tracer of atmospheric processes like resuspension and re-emission from biomass burnings. This allows us to explain 2/3 of the peaks observed over the last six years. The remaining 1/3 is mainly noticed in winter when spreading of pollutants in the atmosphere is often weak due to temperature gradient inversion. On average, continental air masses are responsible for increases by a factor of 3 while oceanic air masses are characterised by levels 3 times lower, compared to the mean value. Feeding of 137Cs in air at ground-level is the result of both local resuspension that signs 137Cs activity levels in soils to which is added a remote contribution from time to time (resuspension of Saharan dust or re-emission from fires occurring in eastern territories with high 137Cs deposition level). Finally, 137Cs activity levels in air masses crossing over France can be described on average by a longitudinal gradient.