Valentin Tanasa

Backstepping Control Under Multi-Rate Sampling

The paper deals with the design of sampled-data controllers which preserve the stabilizing performance of a continuous-time backstepping control strategy. This is achieved through matching of the control Lyapunov functions evolutions at the sampling instants. The method is developed for systems in strict-feedback form. The results are discussed and compared with similar strategies through simulated examples.

Experimental Validation of a Sampled-Data Passivity-Based Controller for Coordination of Converters in a Fuel Cell System

This paper deals with the control of a hybrid source combining a proton exchange membrane fuel cell and supercapacitors. To achieve this objective, an interconnection and damping assignment passivity-based control in a sampled-data context is implemented. This paper first details with this new controller and shows that it ensures the stabilization of the hybrid system at its desired equilibrium point under large range of sampling periods. These considerations are followed by a detailed discussion on experimental results achieved on a 1.2-kW test bench.

Digital stabilization of delayed-input strict-feedforward dynamics

The paper deals with sampled-data stabilization of delayed input-affine dynamics. To remove the infinite dimensionality, the design is set in the sampled-data context and the delay is compensated with a discrete-time predictor. A simulated example illustrates the performances.

A sampled-data level control of nonlinear coupled-tanks

The object of this work is to highlight, on a physical coupled tanks system, the performances of a digital design of a continuous-time passivity based controller. The design considers the sampled-data model of the plant and provides better results compared to the emulations of given continuous-time controllers. A two-state continuous-time non-linear model is used to design the digital controller. A feedback passivating controller is developed in continuous-time from a suitable choice of the storage function V. The results show good performances of the controller when increasing the sampling period or the controllers gain.

Nonlinear optimal stabilizing control under sampling

Starting from an optimal stabilizing controller designed for an input-affine continuous-time dynamics, the paper discusses the problem of computing a digital controller which preserves these properties under sampling. A direct digital approach which minimizes a modified cost functional is described. Some simulated examples illustrate the performances.

Experimental digital control of a magnetic suspension

In this paper, the performances of a digital back-stepping stabilizing control strategy are tested on a real plant. This method applies to continuous-time systems in strict feedback forms which admit a stabilizing backstepping controller. Simulation and experimental results illustrate the benefits of the proposed digital strategy with respect to emulated solutions. References tracking is achieved with less control effort even when considering large sampling periods.

A Computer Aided Software for Nonlinear Digital Control

The aim of this paper is to provide a basic software tool involved in the design and the computation of nonlinear digital control schemes. The complete software SimNLSys will include different modules associated with each control scheme, aimed with respective objectives such as stabilization, damping, output matching or optimality, to quote a few. The software module presented here also encompasses the goal of computing solutions to nonlinear ODEs describing input-affine dynamics. The underlying methodologies propose challenging nonlinear sampled-data control strategies described in terms of Lie series expansions and computable through formal series calculus. The goal is to reach a friendly graphical user interface, to promote the practical efficiency of such strategies. For this, SymNLSys is built using the symbolic Matlab toolbox and the resulting mathematical control law expressions can be implemented directly on experimental plants. The paper describes the used differential geometric methods and discusses the implementation and limitations of the symbolic and numerical algorithms. Also some results obtained via this software are provided.