Bayesian inference and learning in Gaussian process state-space models with Particle MCMC

State-space models are successfully used in many areas of science, engineering and economics to model time series and dynamical systems. We present a fully Bayesian approach to inference and learning (i.e. state estimation and system identification) in nonlinear nonparametric state-space models. We place a Gaussian process prior over the state transition dynamics, resulting in a flexible model able to capture complex dynamical phenomena. To enable efficient inference, we marginalize over the transition dynamics function and, instead, infer directly the joint smoothing distribution using specially tailored Particle Markov Chain Monte Carlo samplers. Once a sample from the smoothing distribution is computed, the state transition predictive distribution can be formulated analytically. Our approach preserves the full nonparametric expressivity of the model and can make use of sparse Gaussian processes to greatly reduce computational complexity.

Quantitative relationships between molecular structures, environmental temperatures and octanol-air partition coefficients of PCDD/Fs

A concise quantitative model that incorporates information on both environmental temperature M and molecular structures, for logarithm of octanol-air partition coefficient (K-OA) to base 10 (logK(OA)) of PCDDs, was developed. Partial least squares (PLS) analysis together with 14 quantum chemical descriptors were used to develop the quantitative relationships between structures, environmental temperatures and properties (QRSETP) model. It has been validated that the obtained QRSETP model can be used to predict logK(OA) of other PCDDs. Molecular size, environmental temperature (T), q(+) (the most positive net atomic charge on hydrogen or chlorine atoms in PCDD molecules) and E-LUMO (the energy of the lowest unoccupied molecular orbital) are main factors governing logK(OA) of PCDD/Fs under study. The intermolecular dispersive interactions and thus the size of the molecules play a leading role in governing logK(OA). The more chlorines in PCDD molecules, the greater the logK(OA) values. Increasing E-LUMO values of the molecules leads to decreasing logK(OA) values, implying possible intermolecular interactions between the molecules under study and octanol molecules. Greater q(+) values results in greater intermolecular electrostatic repulsive interactions between PCDD and octanol molecules and smaller logK(OA) values. (C) 2002 Elsevier Science B.V. All rights reserved.

Crystallization of amorphous Si films by pulsed laser annealing and their structural characteristics

Nanocrystalline silicon (nc-Si) films were prepared by pulsed laser annealed crystallization of amorphous silicon (alpha-Si) films on SiO2-coated quartz or glass substrates. The effect of laser energy density on structural characteristics of nc-Si films was investigated. The Ni-induced crystallization of the a-Si films was also discussed. The surface morphology and microstructure of these films were characterized by scanning electron microscopy, high-resolution electron microscopy, atomic force microscopy and Raman scattering spectroscopy. The results show that not only can the alpha-Si films be crystallized by the laser annealing technique, but also the size of Si nanocrystallites can be controlled by varying the laser energy density. Their average size is about 4-6 nm. We present a surface tension and interface strain model used for describing the laser annealed crystallization of the alpha-Si films. The doping of Ni atoms may effectively reduce the threshold value of laser energy density to crystallize the alpha-Si films, and the flocculent-like Si nanostructures could be formed by Ni-induced crystallization of the alpha-Si films.

Hydrate dissociation conditions for gas mixtures containing carbon dioxide, hydrogen, hydrogen sulfide, nitrogen, and hydrocarbons using SAFT

A new method, a molecular thermodynamic model based on statistical mechanics, is employed to predict the hydrate dissociation conditions for binary gas mixtures with carbon dioxide, hydrogen, hydrogen sulfide, nitrogen, and hydrocarbons in the presence of aqueous solutions. The statistical associating fluid theory (SAFT) equation of state is employed to characterize the vapor and liquid phases and the statistical model of van der Waals and Platteeuw for the hydrate phase. The predictions of the proposed model were found to be in satisfactory to excellent agreement with the experimental data.

Investigation of high cycle and Very-High-Cycle Fatigue behaviors for a structural steel with smooth and notched specimens

The high cycle and Very-High-Cycle Fatigue (VHCF) properties of a structural steel with smooth and notched specimens were studied by employing a rotary bending machine with frequency of 52.5 Hz. For smooth specimens, VHCF failure did occur at fatigue cycles of 7.1 x 10(8) with the related S-N curve of stepwise tendency. Scanning Electron Microscopy (SEM) was used for the observations of the fracture surfaces It shows that for smooth specimens the crack origination is surface mode in the failure regime of less than 10(7) cycles While at VHCF regime, the material failed from the nonmetallic inclusion lies in the interior of material, leading to the formation of fisheye pattern. The dimensions of crack initiation region were measured and discussed with respect to the number of cycles to failure. The mechanism analysis by means of low temperature fracture technique shows that the nonmetallic inclusion in the interior of specimen tends to debond from surrounding matrix and form a crack. The crack propagates and results to the final failure. The stress intensity factor and fatigue strength were calculated to investigate the crack initiation properties. VHCF study on the notched specimens shows that the obtained S-N curve decreases continuously. SEM analysis reveals that multiple crack origins are dominant on specimen surface and that fatigue crack tends to initiate from the surface of the specimen. Based on the fatigue tests and observations, a model of crack initiation was used to describe the transition of fatigue initiation site from subsurface to surface for smooth and notched specimens. The model reveals the influences of load, grain size, inclusion size and surface notch on the crack initiation transition. (C) 2010 Elsevier Ltd. All rights reserved

Performance of supersonic model combustors with staged injections of supercritical aviation kerosene

Supersonic model combustors using two-stage injections of supercritical kerosene were experimentally investigated in both Mach 2.5 and 3.0 model combustors with stagnation temperatures of approximately 1,750 K. Supercritical kerosene of approximately 760 K was prepared and injected in the overall equivalence ratio range of 0.5-1.46. Two pairs of integrated injector/flameholder cavity modules in tandem were used to facilitate fuel-air mixing and stable combustion. For single-stage fuel injection at an upstream location, it was found that the boundary layer separation could propagate into the isolator with increasing fuel equivalence ratio due to excessive local heat release, which in turns changed the entry airflow conditions. Moving the fuel injection to a further downstream location could alleviate the problem, while it would result in a decrease in combustion efficiency due to shorter fuel residence time. With two-stage fuel injections the overall combustor performance was shown to be improved and kerosene injections at fuel rich conditions could be reached without the upstream propagation of the boundary layer separation into the isolator. Furthermore, effects of the entry Mach number and pilot hydrogen on combustion performance were also studied.

Reduction-Based Model Updating of a Scaled Offshore Platform Structure

This paper attempts to develop a reduction-based model updating technique for jacket offshore platform structure. A reduced model is used instead of the direct finite-element model of the real structure in order to circumvent such difficulties as huge degrees of freedom and incomplete experimental data that are usually civil engineers' trouble during the model updating. The whole process consists of three steps: reduction of FE model, the first model updating to minimize the reduction error, and the second model updating to minimize the modeling error of the reduced model and the real structure. According to the performance of jacket platforms, a local-rigidity assumption is employed to obtain the reduced model. The technique is applied in a downscale model of a four-legged offshore platform where its effectiveness is well proven. Furthermore, a comparison between the real structure and its numerical models in the following model validation shows that the updated models have good approximation to the real structure. Besides, some difficulties in the field of model updating are also discussed.

alpha-decay half-lives of superheavy nuclei and general predictions

The generalized liquid drop model (GLDM) and the cluster model have been employed to calculate the alpha-decay half-lives of superheavy nuclei (SHN) using the experimental alpha-decay Q values. The results of the cluster model are slightly poorer than those from the GLDM if experimental Q values are used. The prediction powers of these two models with theoretical Q values from Audi et al. (Q(Audi)) and Muntian et al. (Q(M)) have been tested to find that the cluster model with Q(Audi) and Q(M) could provide reliable results for Z > 112 but the GLDM with Q(Audi) for Z <= 112. The half-lives of some still unknown nuclei are predicted by these two models and these results may be useful for future experimental assignment and identification.

Transfer ionization of helium in collisions with Cq+ and Oq+ (q=1-3)

Cross-section ratios sigma(TI)/sigma(SC) of transfer ionization (TI) to single capture (SC) of Cq+- and Oq+-He (q = 1 - 3) collisions in the energy range of 15-440 keV/u (0.8-4.2 v(Bohr)) are experimentally determined. It is shown that sigma(TI)/sigma(SC) strongly depends on the projectile velocity, and there is a maximum for E(keV/u)/q(1/2) approximate to 150. Combining the Bohr-Lindhard model and the statistical model, a theoretical estimate is presented, in reasonable agreement with the experimental data when E(keV/u)/q(1/2) > 35.

Neutron spectroscopic factors of Ar-34 and Ar-46 from (p,d) transfer reactions

Single-neutron-transfer measurements using (p,d) reactions have been performed at 33 MeV per nucleon with proton-rich Ar-34 and neutron-rich Ar-46 beams in inverse kinematics. The extracted spectroscopic factors are compared to the large-basis shell-model calculations. Relatively weak quenching of the spectroscopic factors is observed between Ar-34 and Ar-46. The experimental results suggest that neutron correlations have a weak dependence on the asymmetry of the nucleus over this isotopic region. The present results are consistent with the systematics established from extensive studies of spectroscopic factors and dispersive optical-model analyses of Ca40-49 isotopes. They are, however, inconsistent with the trends obtained in knockout-reaction measurements.

Effects of elastic and inelastic NN scattering cross sections on pi(-)/pi(+) ratios in heavy-ion collisions at intermediate energies

Based on the isospin-dependent Boltzmann-Uehling-Uhlenbeck transport model and the scaling model according to nucleon effective mass, effects of elastic and inelastic NN scattering cross sections on pi(-)/pi(+) in the neutron-rich reaction of Ca-48 + Ca-48 at a beam energy of 400 MeV/nucleon are studied. It is found that cross-section effects of both NN elastic and inelastic scatterings affect Delta(1232), pi(-) and pi(+) production, as well as the value of pi(-)/pi(+).

Genomic organization, nucleotide sequence analysis of the core histone genes cluster in Chlamys farreri and molecular evolution assessment of the H2A and H2B

This work represents the nucleotide sequence of the core histone gene cluster from scallop Chlamys farreri. The tandemly repeated unit of 5671 bp containing a copy of the four core histone genes H4, H2B, H2A and H3 was amplified and identified by the techniques of homology cloning and genomic DNA walking. All the histone genes in the cluster had the structures in their 3' flanking region which related to the evolution of histone gene expression patterns throughout the cell cycle, including two different termination signals, the hairpin structure and at least one AATAAA polyadenylation signal. In their 5' region, the transcription initiation sites with a conserved sequence of 5'-PyATTCPu-3' known as the CAP site were present in all genes except to H2B, generally 37-45 bp upstream of the start code. Canonical TATA and CAAT boxes were identified only in certain histone genes. In the case of the promoters of H2B and H2A genes, there was a 5'-GATCC-3' element, which had been found to be essential to start transcription at the appropriate site. After this element, in the promoter of H2B, there was another sequence, 5'-GGATCGAAACGTTC-3', which was similar to the consensus sequence of 5'-GGAATAAACGTATTC-3' corresponding to the H2B-specific promoter element. The presence of enhancer sequences (5'-TGATATATG-3') was identified from the H4 and H3 genes, matching perfectly with the consensus sequence defined for histone genes. There were several slightly more complex repetitive DNA in the intergene regions. The presence of the series of conserved sequences and reiterated sequences was consistent with the view that mollusc histone gene cluster arose by duplicating of an ancestral precursor histone gene, the birth-and-death evolution model with strong purifying selection enabled the histone cluster less variation and more conserved function. Meanwhile, the H2A and the H2B were demonstrated to be potential good marks for phylogenetic analysis. All the results will be contributed to the characterization of repeating histone gene families in molluscs.

Folate-Conjugated Micelles and Their Folate-Receptor-Mediated Endocytosis

A folate-conjugated copolymer PEG-PLA-PLL/folate was synthesized and mixed with pure PEG-PLA-PLL and a fluorescent model drug mFITC to prepare folate-conjugated micelles. The distribution of micelles was studied on cancer-cell-bearing mice via frozen slicing. The results show that mFITC is successfully encapsulated into folate(+) and folate(-)micelles; PEG-PLA-PLL micelles the latter can be internalized by both HeLa and CHO cells without selectivity due to their cationic surface charges, while folate(+)micelles exhibit more preferential endocytosis by HeLa cells than by CHO cells. The folate(-)micelles showed retention in both organs and tumors. The folate(+)micelles are a promising active targeting drug delivery system for FR over-expressing cells and they accumulate in tumor beds.

Fabrication of integrated microelectrodes for electrochemical detection on electrophoresis microchip by electroless deposition and micromolding in capillary technique

A new method for the fabrication of an integrated microelectrode for electrochemical detection (ECD) on an electrophoresis microchip is described. The pattern of the microelectrode was directly made on the surface of a microscope slide through an electroless deposition procedure. The surface of the slide was first selectively coated with a thin layer of sodium silicate through a micromolding in capillary technique provided by a poly(dimethylsiloxane) (PDMS) microchannel; this left a rough patterned area for the anchoring of catalytic particles. A metal layer was deposited on the pattern guided by these catalytic particles and was used as the working electrode. Factors influencing the fabrication procedure were discussed. The whole chip was built by reversibly sealing the slide to another PDMS layer with electrophoresis microchannels at room temperature. This approach eliminates the need of clean room facilities and expensive apparatus such as for vacuum deposition or sputtering and makes it possible to produce patterned electrodes suitable for ECD on microchip under ordinary chemistry laboratory conditions. Also once the micropattern is ready, it allows the researchers to rebuild the electrode in a short period of time when an electrode failure occurs. Copper and gold microelectrodes were fabricated by this technique. Glucose, dopamine, and catechol as model analytes were tested.

NMR Spectra of polyimides: Interpretation and local chain motion study

Three kinds of high-performance polyimides 1 (poly(ketone-imide) PKI), 2 (poly(ether-imide) PEI) and 3 (poly(oxy-imide) POI) were studied using nuclear magnetic resonance (NMR). The NMR spectra of the polyimides were assigned according to the comprehensive consideration of the substitution effect of different substituting groups, viz. distortionless enhancement by polarization transfer (DEPT), no nuclear Overhauser effect (NNE), analysis of relaxation time, and two-dimensional correlated spectroscopy (COSY) techniques. The structural units of these three polyimides were determined. Carbon-13 and proton relaxation times for PEI and PKI were interpreted in terms of segmental motion characterized by the sharp cutoff model of Jones and Stockmayer (JS model) and anisotropic group rotation such as phenyl group rotation and methyl group rotation. Correlation times for the main-chain motion are in the tens of picosecond range which indicates the high flexibility of polyimide chains. Correlation times for phenyl group and methyl group rotations are more than 1 order of magnitude lower and approximately 1 order of magnitude higher than that of the main chain, respectively.