A two-stage algorithm for identification of nonlinear dynamic systems

This paper investigates the two-stage stepwise identification for a class of nonlinear dynamic systems that can be described by linear-in-the-parameters models, and the model has to be built from a very large pool of basis functions or model terms. The main objective is to improve the compactness of the model that is obtained by the forward stepwise methods, while retaining the computational efficiency. The proposed algorithm first generates an initial model using a forward stepwise procedure. The significance of each selected term is then reviewed at the second stage and all insignificant ones are replaced, resulting in an optimised compact model with significantly improved performance. The main contribution of this paper is that these two stages are performed within a well-defined regression context, leading to significantly reduced computational complexity. The efficiency of the algorithm is confirmed by the computational complexity analysis, and its effectiveness is demonstrated by the simulation results.

Making 'party identification' more versatile: operationalising the concept for the multiparty setting

The conventional operationalisation of the concept of party identification is not appropriate for the multiparty setting. I offer new measures that facilitate multiple, and negative as well as positive, identities. Using survey evidence from Northern Ireland, these new measures are validated in a number of ways and their role in a comprehensive model of voting is illustrated.

CaIIKinterstellar observations towards early-type disc and halo stars, abundances and distances of intermediate- and high-velocity clouds

We present Ca II K (lambda(air) = 3933.661 angstrom) interstellar observations towards 20 early-type stars, to place lower distance limits to intermediate- and high-velocity clouds (IHVCs) in their lines of sight. The spectra are also employed to estimate the Ca abundance in the low-velocity gas towards these objects, when combined with Leiden-Dwingeloo 21-cm HI survey data of spatial resolution 0 degrees.5. Nine of the stars, which lie towards IHVC complexes H, K and gp, were observed with the intermediate dispersion spectrograph on the Isaac Newton Telescope at a resolution R = lambda/Delta lambda of 9000 (similar to 33 km s(-1)) and signal-to-noise ratio (S/N) per pixel of 75-140. A further nine objects were observed with the Utrecht Echelle Spectrograph on the William Herschel Telescope at R = 40 000 (similar to 7.5 km s(-1)) and S/N per pixel of 10-25. Finally, two objects were observed in both Ca II K and Na I D lines using the 2D COUDE on the McDonald 2.7-m telescope at R = 35 000 (similar to 8.5 km s(-1)). The abundance of Ca II K {log(10)(A) = log(10)[N(Ca II K)]-log(10)[N(HI)]} plotted against HI column density for the objects in the current sample with heights above the Galactic plane (z) exceeding 1000 pc is found to obey the Wakker & Mathis (2000) relation. Also, the reduced column density of Ca II K as function of z is consistent with the larger sample taken from Smoker et al. (2003). Higher S/N observations than those previously taken towards HVC complex H stars HD 13256 and HILT 190 reinforce the assertion that this lies at a distance exceeding 4000 pc. No obvious absorption is detected in observations of ALS 10407 and HD 357657 towards IVC complex gp. The latter star has a spectroscopically estimated distance of similar to 2040 pc, although this was derived assuming the star lies on the main sequence and without any reddening correction being applied. Finally, no Ca II K absorption is detected towards two stars along the line of sight to complex K, namely PG 1610+529 and PG 1710+490. The latter is at a distance of similar to 700 pc, hence placing a lower distance limit to this complex, where previously only an upper distance limit of 6800 pc was available.

Dissociative electron attachment to HCCCN

Dissociative electron attachment to cyanoacetylene (propiolonitrile) HCCCN has been observed in the electron energy range 0-12 eV. Negative ions are formed in two main bands with maxima at similar to 1.6 eV (CCCN-) and similar to 5.3 eV (CCCN-, CN-, HCC- and CC-). There are also weaker resonances which lead to dissociative electron attachment to form CN-, HCC- and CC- with a maximum intensity at similar to 8.1 eV and CCCN-, CN- and CC- at similar to 11.2 eV. A trace of CCN- is observed at similar to 9.1 eV. The positions of the main dissociative attachment bands observed are close to positions of pi* resonances recently calculated by Sommerfeld and Knecht. Calculations have also been performed in this work, which confirm the position of the p* orbitals. The electron affinity of the CCCN radical is determined as 4.59 +/- 0.25 eV from the threshold for CCCN- formation at 1.32 +/- 0.15 eV. Dissociative electron attachment to this molecule will act as a source of negative ions in extraterrestrial environments where electrons are present with more than 1.3 eV energy.

Scaling of domain periodicity with thickness measured in BaTiO3 single crystal lamellae and comparison with other ferroics

The focused ion beam microscope has been used to cut parallel-sided {100}-oriented thin lamellae of single crystal barium titanate with controlled thicknesses, ranging from 530 nm to 70 nm. Scanning transmission electron microscopy has been used to examine domain configurations. In all cases, stripe domains were observed with {011}-type domain walls in perovskite unit-cell axes, suggesting 90 degrees domains with polarization in the plane of the lamellae. The domain widths were found to vary as the square root of the lamellar thickness, consistent with Kittel's law, and its later development by Mitsui and Furuichi and by Roytburd. An investigation into the manner in which domain period adapts to thickness gradient was undertaken on both wedge-shaped lamellae and lamellae with discrete terraces. It was found that when the thickness gradient was perpendicular to the domain walls, a continuous change in domain periodicity occurred, but if the thickness gradient was parallel to the domain walls, periodicity changes were accommodated through discrete domain bifurcation. Data were then compared with other work in literature, on both ferroelectric and ferromagnetic systems, from which conclusions on the widespread applicability of Kittel's law in ferroics were made.

High harmonic generation in the relativistic limit

The generation of extremely bright coherent X-ray pulses in the femtosecond and attosecond regime is currently one of the most exciting frontiers of physics - allowing, for the first time, measurements with unprecedented temporal resolution(1-6). Harmonics from laser - solid target interactions have been identified as a means of achieving fields as high as the Schwinger limit(2,7) (E = 1.3 x 10(16) V m(-1)) and as a highly promising route to high-efficiency attosecond (10(-18) s) pulses(8) owing to their intrinsically phase-locked nature. The key steps to attain these goals are achieving high conversion efficiencies and a slow decay of harmonic efficiency to high orders by driving harmonic production to the relativistic limit(1). Here we present the first experimental demonstration of high harmonic generation in the relativistic limit, obtained on the Vulcan Petawatt laser(9). High conversion efficiencies (eta> 10(-6) per harmonic) and bright emission (> 10(22) photons s(-1) mm(-2) mrad(-2) (0.1% bandwidth)) are observed at wavelengths <4 nm ( the 'water-window' region of particular interest for bio-microscopy).

Proton Radiography of a Laser-Driven Implosion

Protons accelerated by a picosecond laser pulse have been used to radiograph a 500 mu m diameter capsule, imploded with 300 J of laser light in 6 symmetrically incident beams of wavelength 1.054 mu m and pulse length 1 ns. Point projection proton backlighting was used to characterize the density gradients at discrete times through the implosion. Asymmetries were diagnosed both during the early and stagnation stages of the implosion. Comparison with analytic scattering theory and simple Monte Carlo simulations were consistent with a 3 +/- 1 g/cm(3) core with diameter 85 +/- 10 mu m. Scaling simulations show that protons > 50 MeV are required to diagnose asymmetry in ignition scale conditions.