Fuzzy Coding of Genetic Algorithms

A well-cited paper suggesting fuzzy coding as an alternative to the conventional binary, grey and floating-point representations used in genetic algorithms.

A Hybrid Forward Algorithm for RBF Neural Network Construction

This paper proposes a novel hybrid forward algorithm (HFA) for the construction of radial basis function (RBF) neural networks with tunable nodes. The main objective is to efficiently and effectively produce a parsimonious RBF neural network that generalizes well. In this study, it is achieved through simultaneous network structure determination and parameter optimization on the continuous parameter space. This is a mixed integer hard problem and the proposed HFA tackles this problem using an integrated analytic framework, leading to significantly improved network performance and reduced memory usage for the network construction. The computational complexity analysis confirms the efficiency of the proposed algorithm, and the simulation results demonstrate its effectiveness

On the electron impact ionization of multi-charged ions: a theoretical investigation of relativistic effects

The electron impact ionization of highly charged ions is investigated. Using a relativistic distorted wave Born approximation, we explore the possible effects that should be observable in a high-energy electron impact ionization (e, 2e) coincidence experiment involving multi-charged ions. We present calculations of triple-differential cross sections. We will focus on relativistic and distortion effects and consider geometries where these may be easily observed.

A comparison of theoretical line intensity ratios for Ni XII with extreme ultraviolet observations from the JET tokamak

Recent R-matrix calculations of electron impact excitation rates in Ni XII are used to derive the emission line ratios R-1 = I(154.17 Angstrom)/I(152.15 Angstrom), R-2 = I(152.95 Angstrom)/I(152.15 Angstrom) and R-3 = 1(160.55 Angstrom)/I(152.15 Angstrom). This is the first time (to our knowledge) that theoretical emission line ratios have been calculated for this ion. The ratios are found to be insensitive to changes in the adopted electron density (N-e) when N-e greater than or equal to 5 x 10(11) cm(-3), typical of laboratory plasmas. However, they do vary with electron temperature (T-e), with for example R-1 and R-3 changing by factors of 1.3 and 1.8, respectively, between T-e = 10(5) and 10(6) K. A comparison of the theoretical line ratios with measurements from the Joint European Tents (JET) tokamak reveals very good agreement between theory and observation for R-1, with an average discrepancy of only 7%. Agreement between the calculated and experimental ratios for R-2 and R-3 is less satisfactory, with average differences of 30 and 33%, respectively. These probably arise from errors in the JET instrument calibration curve. However, the discrepancies are smaller than the uncertainties in the R-2 and R-3 measurements. Our results, in particular for R-1, provide experimental support for the accuracy of the Ni XIII line ratio calculations, and hence for the atomic data adopted in their derivation.

A comparison of theoretical Si VIII emission line ratios with observations from SERTS

Recent R-matrix calculations of electron impact excitation rates in N-like Si VIII are used to derive theoretical emission line intensity ratios involving 2s(2)2p(3)-2s2p(4) transitions in the 216 -320 Angstrom wavelength range. A comparison of these with an extensive dataset of solar active region, quiet- Sun, sub-flare and off-limb observations, obtained during rocket flights of the Solar EUV Research Telescope and Spectrograph (SERTS), indicates that the ratio R-1 = I(216.94 Angstrom)/I(319.84 Angstrom) may provide a usable electron density diagnostic for coronal plasmas. The ratio involves two lines of comparable intensity, and varies by a factor of about 5 over the useful density range of 10(8)-10(11) cm(-3). However R-2 = I(276.85 Angstrom)/I(319.84 Angstrom) and R-3 = I(277.05 Angstrom)/I(319.84 Angstrom) show very poor agreement between theory and observation, due to the severe blending of the 276.85 and 277.05 Angstrom lines with Si VII and Mg VII transitions, respectively, making the ratios unsuitable as density diagnostics. The 314.35 Angstrom feature of Si VIII also appears to be blended, with the other species contributing around 20% to the total line flux.

A comparison of theoretical MgVI emission line strengths with active-region observations from serts

R-matrix calculations of electron impact excitation rates in N-like Mg vi are used to derive theoretical electron-density-sensitive emission line ratios involving 2s(2)2p(3)-2s2p(4) transitions in the 269-403 Angstrom wavelength range. A comparison of these with observations of a solar active region, obtained during the 1989 flight of the Solar EUV Rocket Telescope and Spectrograph (SERTS), reveals good agreement between theory and observation for the 2s(2)2p(3) S-4-2s2p(4) P-4 transitions at 399.28, 400.67, and 403.30 Angstrom, and the 2s(2)2p(3) P-2-2s2p(4) D-2 lines at 387.77 and 387.97 Angstrom. However, intensities for the other lines attributed to Mg vi in this spectrum by various authors do not match the present theoretical predictions. We argue that these discrepancies are not due to errors in the adopted atomic data, as previously suggested, but rather to observational uncertainties or mis-identifications. Some of the features previously identified as Mg vi lines in the SERTS spectrum, such as 291.36 and 293.15 Angstrom, are judged to be noise, while others (including 349.16 Angstrom) appear to be blended.