Simplified robust control for nonlinear uncertain systems: a method of projection and online estimation

Robust control based on an online estimation of uncertainty is presented for a class of nonlinear uncertain systems. The estimation is done via a robust observer after the uncertainty vector is projected onto a one-dimensional subspace. The proposed combination of dynamics projection and online estimation is to relax the knowledge about the size of uncertainty and required in the robust control design, to make robust control less conservative while being effective, and to ensure robust stability without undue complexity.     

First-principles calculations of the elastic,phonon and thermodynamic properties of Al12Mg17

The elastic, phonon and thermodynamic properties of Al₁₂Mg₁₇ have been investigated by first-principles calculations. The obtained structural parameters, phonon dispersion curves and the predicted thermodynamic properties for all the phases studied herein agree well with available experimental data. The temperature-dependent single-crystal elastic constants are also predicted along with the polycrystalline aggregate properties, including bulk modulus, shear modulus, Young’s modulus, and Poisson’s ratio. The brittleness of Al₁₂Mg₁₇ that we predict is consistent with experiments, in contrast to the previous calculation showing ductile behavior. Detailed analysis of density of states further explains the present theoretical findings.     

2-D Shape Recognition using Recursive Landmark Determination and Fuzzy ART Network Learning

In this Letter, 2-D shape recognition is done using a combination of recursive search of landmarks, landmark-based invariant features, and a fuzzy ART neural-network classifier. To make this novel combination work well, an upper limit is imposed on the number of total landmarks allowed, and this maximum size is then translated into fixed dimensions of invariant features and into the neural processing of the features. It is shown that the recursive landmark search approximates very well any smooth 2-D shape contour, that the shape features used are independent of perspective transformation, and that, when combinedwitha fuzzy ART classifier, unknown features can be efficiently learned on-line to identify multiple distinct objects. An illustrative example is used to demonstrate effectiveness of the proposed algorithm.     

Formation of the Face-Centered Cubic (FCC)-NdO x Phase at Nd/Nd-Fe-B Interface: A First-Principles Modeling

The development of advanced Nd-Fe-B permanent magnet materials with high coercivity draws much attention to the relation between coercivity and microstructure at the grain boundaries of the magnets. A disordered face-centered cubic (fcc)-NdO _x phase formed at the interface of Nd/Nd-Fe-B is observed, and it is believed to take an important role in coercivity generation. To have a thorough understanding of the formation mechanism of this particular oxide and its relation to the surface coercivity, a ground state analysis for whole oxygen concentration in Nd-O has been performed by combining the LSDA + U and the cluster expansion method. Systematic calculations revealed that a sequent fcc-based structure formed by introducing oxygen vacancies into NdO is stable in almost all the 0–50% oxygen concentration range, whereas in a series of hexagonal close-packed (hcp)-based structures developed from hP5-Nd₂O₃ no stable structure is observed, which coincides with the experimental measurement very well. A further analysis of formation energies and relevant changes in electronic structures of single oxygen vacancy in various structures revealed the insight of such fcc-based phase formation and further explained the relation between the phase stability and coercivity.     

Defects in boron carbide: First-principles calculations and CALPHAD modeling

The energetics of defects in B_4+xC boron carbide and β-boron are studied through first-principles calculations, the supercell phonon approach and the Debye–Grüneisen model. It is found that suitable sublattice models for β-boron and B_4+xC are B₁₀₁(B,C)₄ and B₁₁(B,C) (B,C,Va) (B,Va) (B,C,Va), respectively. The thermodynamic properties of B_4+xC, β-boron, liquid and graphite are modeled using the CALPHAD approach based on the thermochemical data from first-principles calculations and experimental phase equilibrium data in the literature. The concentrations of various defects are then predicted as a function of carbon composition and temperature.     

Enthalpies of formation of magnesium compounds from first-principles calculations

An energetics database of binary magnesium compounds has been developed from first-principles calculations. The systems investigated include Mg–X (X = As, Ba, Ca, Cd, Cu, Dy, Ga, Ge, La, Lu, Ni, Pb, Sb, Si, Sn and Y). The calculated lattice parameters and enthalpies of formation of binary compounds in these systems are compared with both experimental data and thermodynamic databases.     

Robust fault-tolerant self-recovering control of nonlinear uncertain systems

In this paper, the problem of devising a fault-tolerant robust control for a class of nonlinear uncertain systems is investigated. Possible failures of the sensor measuring the state variables are considered, and a robust measure is developed to identify the stability- and performance-vulnerable failures. Based on evaluation of the robust measure, a fault-tolerant robust control will switch itself between one robust control strategy designed under normal operation and another under the faulty condition. It is shown that, under two input-to-state stability conditions, the proposed scheme guarantees not only the desired performance under normal operations but also robust stability and best achievable performance when there is a sensor failure of any kind.