Search

The development of product-service systems is affected by a large number of influences. These influences often show a cyclic character. The management of these cyclic influences reveals great potential for improving the product-service system development process. Knowledge about influences and their interdependencies supports the prediction of future changes and is highly valuable for planning. A core challenge is the high degree of dependencies between the different cycles.

Reinforcement learning offers a multitude of algorithms allowing to learn a nonlinear controller by interacting with the system without the need for a model of the plant. In this paper we investigate the suitability of online learning algorithms for a control task with incomplete state information. The system under consideration is a swinging chain that needs to be stabilized at a desired position, a problem that is occurring e.g. with bridge cranes with each change in the crane position. The measurable states are the position, velocity, angle and angular velocity at the top of the chain. A solution of the control problem based on an approximation of the chain as a continuous cable exists in the literature, see d’ Andrea-Novel and Coron (2000), which is included in the comparison as a reference for the control performance of the learned controllers.

AbstractThis paper considers the problem of complexity in reduced nonlinear models derived by System Matrices Optimization method. The solution suggested here is to add a class of secondary conditions that impose a simpler structure on the reduced model. Distinct choices of these conditions have different impacts on the accuracy of the reduced model, therefore a pioneered search based on genetic algorithms is proposed that finds the optimal choice of conditions. By means of two examples, it is illustrated that this method can effectively compromise between simplicity and accuracy of the reduced model. In order to improve the numerical efficiency of the solution and to speed up the pace of convergence some enhancements are presented.

A method is proposed for calculating alignment and orientation for open shell atoms after photoionization of an inner shell by relating the open shell dipole matrix elements to those independent of the outer open shell configuration. Thus, an extended and usually non-trivial numerical open shell approach is avoided. Our method can be applied to different configuration potentials which is illustrated by Hartree-Fock calculations of alignment and orientation which have been performed for the L3photoionization of K, Au, Sc and Mn, and for the M5and N7photoionization of In and Tl, applying different numerical approaches. Dramatic effects have been found for Mn. The alignment of Au is compared to the experiment.

A method is proposed for calculating alignment and orientation for open shell atoms after photoionization of an inner shell by relating the open shell dipole matrix elements to those independent of the outer open shell configuration. Thus, an extended and usually non-trivial numerical open shell approach is avoided. Our method can be applied to different configuration potentials which is illustrated by Hartree–Fock calculations of alignment and orientation which have been performed for the L3 photoionization of K, Au, Sc and Mn, and for the M5 and N7 photoionization of In and Tl, applying different numerical approaches. Dramatic effects have been found for Mn. The alignment of Au is compared to the experiment.

Experimental and theoretical results are presented for the A2spin asymmetry for elastic and inelastic scattering of spin-polarized electrons from rubidium, at incident electron energies of 15, 20, 30, 50 and 80 eV. Theoretical calculations are performed within a semi-relativistic Breit-Pauli R-matrix (close-coupling) model. The experimental data reveal significant asymmetries at all energies. The experimental results are reasonably well described by the present theoretical calculations and by other calculations using a relativistic distorted-wave approach, and provide a clear relativistic signature.

Abstract. In this paper, a Haar wavelet-based method for optimal control of the second-order linear systems with respect to a quadratic cost function for any length of time is proposed. A Haar wavelet integral operational matrix and properties of the Kronecker product are used in finding the approximate solutions of optimal trajectories and optimal control by solving only two algebraic equations instead of solving the Riccati differential equation. Numerical results of a typical example are presented to illustrate the advantage of the approach.

This paper presents an application of wavelet networks in identification and control design for a class of non-linear dynamical systems. The technique of feedback linearization, supervisory control and H∞control are used to design an adaptive control law and also the parameters of the wavelet networks are adapted using a Lyapunov-based design. By some theorems, it will be proved that even in the presence of modelling errors, named network error, the stability of the overall identification scheme and the convergence of the network parameters and the boundedness of the state errors are guaranteed.

The problem of Lyapunov Exponents (LEs) estimation from chaotic data based on Jacobian approach by polynomial models is considered. The optimum embedding dimension of reconstructed attractor is interpreted as suitable order of model. Therefore, based on global polynomial mode ling of system, a novel criterion for selecting the embedding dimension is presented. By considering this dimension as the model order, the best nonlinearity degree of polynomial is estimated. The selected structure is used for local estimating of Jacobians to calculate the LEs. This suitable structure of polynomial model leads to better results and avoids of sporious LEs. Simulation results show the effectiveness of proposed methodology.

AbstractEin neuartiger visueller Regler zur autonomen, bildbasierten Navigation einer mobilen Plattform (Rollstuhl) wird vorgestellt. Ziel ist es, den Rollstuhl mittels Image-Based Visual Servoing in eine relative Lage bezüglich einer Landmarke zu navigieren, was allein durch Auswertung von Abweichungen im Kamerabild und ohne Bestimmung der Landmarkenposition in Weltkoordinaten geschieht. Dies wird mithilfe des in diesem Artikel vorgestellten 3D-Visual Servoing erreicht, das eine Erweiterung des herkömmlichen 2D-Visual Servoing darstellt, indem die Entfernung zwischen Kamera und Landmarke mit in die bildbasierte Regelung einbezogen wird. Der visuelle Regler wird dabei durch eine unterlagerte Trajektorienfolgeregelung unterstützt, welche die Fahrt der Plattform überwacht. Zusätzlich wird die Kamera drehbar auf einem Schwenk-Neige-Kopf gelagert, sodass die Landmarke während der Fahrt im Kamerabild gehalten werden kann.

Tight time and cost constraints force chemical process industries to continuously reduce their experimental effort during chemical process development and to facilitate the application of model-based process technology. Nevertheless, the effort of setting up a detailed mathematical model for a chemical process remains still high due to the large variety of chemical process units and physical phenomena as well as increasing requirements on the sophistication of models. To overcome this modeling bottleneck considerable effort has to be spent with the systematization of process models, the formalization of their representation, and the development of knowledge-based software tools. This contribution deals with the modeling environment ModKit, which supports the systematic development, maintenance, and reuse of chemical process models. It will be shown, how informal textual or graphical information defining requirements on models can be specified by use of ModKit. Furthermore, computer-aided means for the structural and behavioural description of chemical processes, as well as an approach for the support of the workflow of model development will be described.

AbstractIn this article a high-gain decentralized controller is designed for a large-scale system. The effects of the interactions between the subsystems are considered as uncertainty for the large-scale system. A bound on the high-gain factor is computed to nullify the effects of the interactions and also to ensure the overall closed-loop stability. In order to avoid saturation, the anti-windup integrator method is used in designing high-gain controller. Due to high-gain feedback, the closed-loop system is robust with respect to output disturbances and uncertainties.

This paper analyzes the modeling process by means of a field study in the chemical industries. A diversified set of experienced modelers from two different countries have been interviewed using a case study approach. The interviews focussed on the modeling process. Other issues like life-cycle perspectives and the use of modeling in the process industries were treated as well. The information has been summarized and presented in a structured manner imposing the least possible bias from the authors. Based on the interview information the modeling process is discussed in detail. Further, fundamental research issues are identified, and a research agenda is proposed.

The decoupling of the input - output behaviour of linear multivariable systems by constantstate- feedback generally requires the compensation of all invariant zeros, which causes unstability in the case of non - minimum phase systems. The paper presents a method for partial but stable decoupling, ensuring without compensations coupling of only oneoutput with several inputs. The application of a dynamicstate-feedback however allows complete and stable decoupling with the disadvantage of an increased number of disagreeable transmission zeros.

The conjugational transfer of the plasmid pAM β1 between two strains of Lactococcus lactis subsp. lactis was measured in the intestinal tract of four conventional and eight gnotobiotic rats. In the conventional rats both donor and recipient strains were eliminated and transconjugants were not observed. Germfree rats were dosed orally with the recipient strain and the donor strain carrying pAM β1. In faecal samples, transconjugants were detected within the first few days after dosing, and the transconjugants established a stable population at a level of 103 - 10s less than the recipient strain. In one group of animals the donor strain was eliminated from the intestine shortly after introduction, but still the transconjugants colonised the intestine. Samples from duodenum, jejunum, caecum and colon were cultured 2 or 5 wks after dosing with the recipient strain. The concentration of transconjugants was approximately 104c.f.u./g throughout the intestine, whereas the concentration of recipients increased from 104-105 c.f.u./g in jejunum to 108-109 c.f.u./g in caecum and colon. The plasmid pAMβ1 seems to have conveyed to the recipient a competitive advantage in the small intestine but not in the large intestine.

A new sufficient condition is presented for the overall stability of decentralised linear control systems. This condition is in terms of the eigenvalues of the Hermitian part of the interaction matrix and the Hermitian part of the state matrix of each closed-loop isolated subsystem.

A black-box method for parametric model order reduction is presented that includes method selection, model refinement and error prediction using a cross-validation-based error indicator. The method is demonstrated for the interpolation of reduced system matrices.

The presented approach allows a considerable reduction of the system order of medium and large-scale systems without loss of the characteristic system behaviour. Such order-reductions are useful when nonlinear feedback design or fast simulations require a low system order.

Stress roentgenograms in 25 probands were performed to evaluate the stability of the normal shoulder joint. The shoulder was remarkably stable, the average ventral subluxation being 0.8 mm, the dorsal subluxation 0.2 mm.

The angular distribution and spin polarisation of the resonantly photoexcited Xe¤(4d–15/2 6p3/2 ) N5O2,3O2,3 Auger spectrum is investigated. The two-step model has been used which allows us to independently determine the dynamic parameters of the primary excitation and the Auger emission process. Assuming either a fully circularly or linearly polarised photon beam the dynamic parameters determining the primary photoexcitation become constant numbers independent of the matrix elements. Applying a relativistic distorted wave approximation the relevant numbers describing the Auger decay dynamics, i.e. relative intensities, angular distribution and spin polarisation parameters have been calculated, and are compared with experimental and other theoretical data. With this, predictions for the spin polarisation vector are possible. A large degree of dynamic spin polarisation has been found for all Auger transitions to a final state with Jf = ½ . This is in contrast to earlier calculations for diagram Auger transitions. Recently, we have given an explanation for this deriving propensity rules for resonant Auger transitions. The propensity rules allow for predictions for which Auger line a large dynamic spin polarisation can be expected. The predictions are in accord with our multiconfigurational Dirac–Fock calculations for the resonant Xe N5O2,3O2,3 and Ar L3M2,3M2,3 Auger multiplets. It is demonstrated that the effect of a large spin polarisation is caused by a large shift of the scattering phase of the emitted εs1/2 partial waves, whereas a small spin polarisation is due to a cancellation between the Coulomb and scattering phases of the partial waves.