Research old

 Identification and Control of Linear Systems

Owing to the well-developed theory for linear systems, much of the research in the identification and control field has focused on the development and use of linear models. The fact that real processes are nonlinear and/or time-varying in nature is dealt with by either imposing nonlinear robustness properties to the closed-loop system or adapting the controller via iterative algorithms based on real-time data. The research in the area of linear systems is limited to single-input single-output systems and focus along the following directions: Control-Relevant Identification, Data-Driven Iterative Controller Tuning, Model-Based Control Design Using the Polynomial Approach

Keywords : Linear; Control; 

Optimization has received more attention in industry recently since it is a natural choice for reducing production costs and improving product quality and reproducibility without violating safety requirements and environmental regulations. The main bottleneck in applying standard optimization techniques at the industrial level is the fact that they rely heavily on an accurate mathematical model of the process that, however, is seldom available. Research at the Laboratoire d’Automatique studies various ways of using measurements in the optimization framework to combat the lack of process knowlege and process variations. Successful simulation studies with this approach have been carried out on several batch reactors, biotechnological systems, distillation columns, and polymerization processes. These studies show performance improvement in the range of 20-30 % compared to a conservative strategy. The methodology has also been tested on some laboratory-scale setups such as a batch reactor with safety constraints, a biofilter for waste-water treatment, and a fed-batch fermenter growing Baker’s yeast. Also, an application of this strategy for electro-erosion at an industrial level is underway.

Keywords : Optimization; Measurement-based Optimization; On-line Optimization; Run-to-Run Optimization;

The research in the field of nonlinear control attempts to bridge the gap between advanced theory and concrete applications. On the methodological side, interest is mainly geared towards differential geometry with links to predictive control and nonlinear stabilization using motion planning and tracking. Classical analysis tools such as Lyapunov theory and singular perturbations are used in the analysis phases. The range of applications is broad and includes: Mechanical systems such as hydraulic presses, underactuated mechanical structures for the space industry and piezo-electric actuators. Energy systems such as a tokamak fusion reactor and a solar water heater for domestic use. Various testbeds such as a toycopter, a pendubot, a crane, a piezo-actuator, nonholonomic robots and solar-heated systems.

Keywords : Feedback linearization; Differential Flatness; Lyapunov Stability; Algebra; Differential Geometry; UAVs; Robotics; Crane Control; Biomechanics; 

This interdisciplinary research aims at handling coordination and interaction challenges using systems engineering methodologies. The underlying approach is to exploit in an integrated manner the dynamical couplings that exit between the users, the physical systems as well as the communication and information processing layers to improve the quality of services in real-time coordination and the quality of experience in distributed interaction. Innovative domains of deployment are telematics, teleoperation of mechatronic systems, autonomous transportation systems and e-Learning.

Keywords : Real-time; interaction; coordination; collaboration; HCI; CSCW; Traffic; Autonomous Vehicle; Quality of Perception; Quality of Service; 

This research aims at the development of process chemometric methods for both the monitoring of not directly measurable variables and the modeling of chemical reac-tion systems in the presence of significant noise and drift. The monitoring of product quality uses spectroscopic or image measurements. The modeling of homogeneous and heterogeneous chemical and biotechnological reaction systems is based on multi-sensor data that include spectroscopic, flowrate, temperature, pressure and calorimet-ric measurements. These methods are applied to various practical problems such as the automatic detection of the pathological state of rabbit kidneys based on magnetic resonance imaging data, the improved prediction of metabolite concentrations in fer-mentations based on infrared measurements, and the kinetic modeling based on infra-red data and flowrates from microreactors.

Keywords : Data analysis; modeling; gas-liquid reaction systems; homogeneous reaction systems; spectroscopic data; calorimetric data; multi-sensor data;