Surrogate-Based Optimization Using Multifidelity Models with Variable Parameterization and Corrected Space Mapping

Surrogate-based-optimization methods provide a means to achieve high-fidelity design optimization at reduced computational cost by using a high-fidelity model in combination with lower-fidelity models that are less expensive to evaluate. This paper presents a provably convergent trust-region model-management methodology for variableparameterization design models: that is, models for which the design parameters are defined over different spaces. Corrected space mapping is introduced as a method to map between the variable-parameterization design spaces. It is then used with a sequential-quadratic-programming-like trust-region method for two aerospace-related design optimization problems. Results for a wing design problem and a flapping-flight problem show that the method outperforms direct optimization in the high-fidelity space. On the wing design problem, the new method achieves 76% savings in high-fidelity function calls. On a bat-flight design problem, it achieves approximately 45% time savings, although it converges to a different local minimum than did the benchmark.

GD 552: a cataclysmic variable with a brown dwarf companion?

GD 552 is a high proper motion star with the strong, double-peaked emission lines characteristic of the dwarf nova class of cataclysmic variable (CV) star, and yet no outburst has been detected during the past 12yr of monitoring. We present spectroscopy taken with the aim of detecting emission from the mass donor in this system. We fail to do so at a level which allows us to rule out the presence of a near-main-sequence star donor. Given GD 552's orbital period of 103 min, this suggests that it is either a system that has evolved through the ~80-minute orbital period minimum of CV stars and now has a brown dwarf mass donor, or that has formed with a brown dwarf donor in the first place. This model explains the low observed orbital velocity of the white dwarf and GD 552's low luminosity. It is also consistent with the absence of outbursts from the system.

SDSS J150722.30+523039.8: a cataclysmic variable formed directly from a detached white dwarf/brown dwarf binary?

We present high-speed, three-colour photometry of the eclipsing cataclysmic variable SDSS J150722.30+523039.8 (hereafter SDSS J1507). This system has an orbital period of 66.61 min, placing it below the observed `period minimum' for cataclysmic variables. We determine the system parameters via a parametrized model of the eclipse fitted to the observed lightcurve by ?2 minimization. We obtain a mass ratio of q = 0.0623 +/- 0.0007 and an orbital inclination . The primary mass is Mw = 0.90 +/- 0.01Msolar. The secondary mass and radius are found to be Mr = 0.056 +/- 0.001Msolar and Rr = 0.096 +/- 0.001Rsolar, respectively. We find a distance to the system of 160 +/- 10pc. The secondary star in SDSS J1507 has a mass substantially below the hydrogen burning limit, making it the second confirmed substellar donor in a cataclysmic variable. The very short orbital period of SDSS J1507 is readily explained if the secondary star is nuclearly evolved, or if SDSS J1507 formed directly from a detached white dwarf/brown dwarf binary. Given the lack of any visible contribution from the secondary star, the very low secondary mass and the low HeI ?6678/Ha emission-line ratio, we argue that SDSS J1507 probably formed directly from a detached white dwarf/brown dwarf binary. If confirmed, SDSS J1507 will be the first such system identified. The implications for binary star evolution, the brown dwarf desert and the common envelope phase are discussed.

The component masses of the cataclysmic variable V347 Puppis

We present time-resolved spectroscopy and photometry of the double-lined eclipsing cataclysmic variable V347 Pup ( = LB 1800). There is evidence of irradiation on the inner hemisphere of the secondary star, which we correct for using a model to give a secondary-star radial velocity of K R = 198 +/- 5 km s(-1). The rotational velocity of the secondary star in V347 Pup is found to be v sin i = 131 +/- 5 km s(-1) and the system inclination is i = 84degrees.0 +/- 2degrees.3. From these parameters we obtain masses of M-1 = 0.63 +/- 0.04 M for the white dwarf primary and M-2 = 0.52 +/- 0.06 M for the M0.5V secondary star, giving a mass ratio of q = 0.83 +/- 0.05. On the basis of the component masses, and the spectral type and radius of the secondary star in V347 Pup, we find tentative evidence for an evolved companion. V347 Pup shows many of the characteristics of the SW Sex stars, exhibiting single-peaked emission lines, high-velocity S-wave components and phase-offsets in the radial velocity curve. We find spiral arms in the accretion disc of V347 Pup and measure the disc radius to be close to the maximum allowed in a pressureless disc.

ULTRACAM photometry of the eclipsing cataclysmic variable OU Vir

We present high-speed, three-colour photometry of the faint eclipsing cataclysmic variable OU Vir. For the first time in OU Vir, separate eclipses of the white dwarf and the bright spot have been observed. We use timings of these eclipses to derive a purely photometric model of the system, obtaining a mass ratio of q=0.175+/-0.025, an inclination of i=79.degrees2+/-0.degrees7 and a disc radius of R-d/a=0.2315+/-0.0150. We separate the white dwarf eclipse from the light curve and, by fitting a blackbody spectrum to its flux in each passband, obtain a white dwarf temperature of T=13900+/-600 K and a distance of D=51+/-17 pc. Assuming that the primary obeys the Nauenberg mass-radius relation for white dwarfs and allowing for temperature effects, we also find a primary mass M-w/M-circle dot=0.89+/-0.20, a primary radius R-w/R-circle dot=0.0097+/-0.0031 and an orbital separation a/R-circle dot=0.74+/-0.05.

The convex variable step-size (CVSS) algorithm

This letter introduces the convex variable step-size (CVSS) algorithm. The convexity of the resulting cost function is guaranteed. Simulations presented show that with the proposed algorithm, we obtain similar results, as with the VSS algorithm in initial convergence, while there are potential performance gains when abrupt changes occur.

Convergence and tracking analysis of a variable normalised LMF (XE-NLMF) algorithm

The least-mean-fourth (LMF) algorithm is known for its fast convergence and lower steady state error, especially in sub-Gaussian noise environments. Recent work on normalised versions of the LMF algorithm has further enhanced its stability and performance in both Gaussian and sub-Gaussian noise environments. For example, the recently developed normalised LMF (XE-NLMF) algorithm is normalised by the mixed signal and error powers, and weighted by a fixed mixed-power parameter. Unfortunately, this algorithm depends on the selection of this mixing parameter. In this work, a time-varying mixed-power parameter technique is introduced to overcome this dependency. A convergence analysis, transient analysis, and steady-state behaviour of the proposed algorithm are derived and verified through simulations. An enhancement in performance is obtained through the use of this technique in two different scenarios. Moreover, the tracking analysis of the proposed algorithm is carried out in the presence of two sources of nonstationarities: (1) carrier frequency offset between transmitter and receiver and (2) random variations in the environment. Close agreement between analysis and simulation results is obtained. The results show that, unlike in the stationary case, the steady-state excess mean-square error is not a monotonically increasing function of the step size. (c) 2007 Elsevier B.V. All rights reserved.