Correcting the systematic error of the density functional theory calculation: the alternate combination approach of genetic algorithm and neural network

The alternate combinational approach of genetic algorithm and neural network (AGANN) has been presented to correct the systematic error of the density functional theory (DFT) calculation. It treats the DFT as a black box and models the error through external statistical information. As a demonstration, the AGANN method has been applied in the correction of the lattice energies from the DFT calculation for 72 metal halides and hydrides. Through the AGANN correction, the mean absolute value of the relative errors of the calculated lattice energies to the experimental values decreases from 4.93% to 1.20% in the testing set. For comparison, the neural network approach reduces the mean value to 2.56%. And for the common combinational approach of genetic algorithm and neural network, the value drops to 2.15%. The multiple linear regression method almost has no correction effect here.

Correlation analysis of the structures and gas-liquid chromatographic retention indices of amines

Quantitative structure-retention relationship(QSRR) was studied for amines to gas-liquid chromatography on three stationary phases of different polarities with the topological indices A(m) (A(m1), A(m2), A(m3)) and gravitational index GI. The algorithm of "Leaps and Bounds" was performed for selection of the variables. And the multi-regression and the quasi-Newton neural networks were employed for the calculation with better results.

Growing season ecosystem respirations and associated component fluxes in two alpine meadows on the Tibetan Plateau

From 30 June to 24 September in 2003 ecosystem respiration (Re) in two alpine meadows on the Tibetan Plateau were measured using static chamber- and gas chromatography- (GC) based techniques. Simultaneously, plant removal treatments were set to partition Re into plant autotrophic respiration (Ra) and microbial heterotrophic respiration (Rh). Results indicated that Re had clear diurnal and seasonal variation patterns in both of the meadows. The seasonal variability of Re at both meadow sites was caused mainly by changes in Ra, rather than Rh. Moreover, at the Kobresia humilis meadow site (K_site), Ra and Rh accounted for 54% and 46% of Re, respectively. While at the Potentilla fruticosa scrub meadow (P_site), the counterparts accounted for 61% and 39%, respectively. T test showed that there was significant difference in Re rates between the two meadows (t = 2.387, P = 0.022). However, no significant difference was found in Rh rates, whereas a significant difference was observed in Ra rates between the two meadows. Thus, the difference in Re rate between the two meadows was mainly attributed to plant autotrophic respirations. During the growing season, the two meadows showed relatively low Q(10) values, suggesting that Re, especially Rh was not sensitive to temperature variation in the growing season. Additionally, Re and Rh at the K_site, as well as Rh at the P_site was negatively correlated with soil moisture, indicating that soil moisture would also play an important role in respirations.

Rarefied Gas Flows in Micro-Channels

利用基于分子模型的统计模拟方法--信息保存方法(IP)统计模拟了实验条件下微槽道气体流动,仔细讨论了用IP方法模拟长槽道稀薄气流时遇到的问题,并给出了解决的方法,即采取守恒形式的控制方程避免质量流量计算误差积累,并利用超松弛方法使收敛过程加速。将IP计算结果与压力分布和质量流量实验数据进行了比较。

Parameters Inversion of Fluid-Saturated Porous Media Based on Neural Networks

The multi-layers feedforward neural network is used for inversion of material constants of fluid-saturated porous media. The direct analysis of fluid-saturated porous media is carried out with the boundary element method. The dynamic displacement responses obtained from direct analysis for prescribed material parameters constitute the sample sets training neural network. By virtue of the effective L-M training algorithm and the Tikhonov regularization method as well as the GCV method for an appropriate selection of regularization parameter, the inverse mapping from dynamic displacement responses to material constants is performed. Numerical examples demonstrate the validity of the neural network method.

Application of Electrical Resistance Tomography System to Monitor Gas/Liquid Two-Phase Flow in a Horizontal Pipe

The Electrical Resistance Tomography (ERT) technique possesses great potential in monitoring widely exiting industrial two/multi-phase flow. For vertical pipe flow and inclined pipe flow, some application studies with exciting results have been reported, but there is rarely a paper regarding the application of ERT to horizontal gas/liquid pipe flow. This paper addresses this issue and proposes a smart method, Liquid Level Detection method, to conventional ERT system. The enhanced ERT system using the new method can monitor horizontal pipe flow effectively and its application is no longer restricted by the flow conditions. Some experimental results from monitoring an air/water slug pipe flow are presented.

Characteristics of Micronozzle Gas Flows

The fluid characteristics of gas flows in the micronozzle whose throat height is 20 μm were investigated by the direct simulation Monte Carlo (DSMC) method. In a series of cases, the dependence of mass flux on the pressure difference was gained, and the DSMC's results show good agreement with the experimental data. The comparison of mass flux and the Mach number contours between the DSMC and Navier-Stokes equations adding slip boundary also reveals quantitatively that the continuum model will be invalid gradually even when the average Knudsen number is smaller than 0.01. As one focus of the present paper, the phenomenon of the multiple expansion-compression waves that comes from the nozzle's divergent part was analyzed in detailed.

Study of Gas Combustion in a Vented Cylindrical Vessel

Complicated interaction of a flame front with a turbulent flow induced by venting is studied during combustion of the stoichiometric propane/air mixture in a relatively large vented cylindrical vessel. Flame position, its shape, and combustion pressure were measured as a function of time and vent parameters. The experimental data were used to verify numerical simulation of the combustion process. The proposed numerical model satisfactorily simulates the main features of combustion in a closed and vented vessel such as flame configuration, flow and temperature fields, and pressure variation pattern. Simulated velocity and temperature distribution are very useful pieces of information because they are not available from experiments.

Experimental study and simulation principles of an oil-gas multiphase transportation system

Presented is an experimental study on the performance of an oil-gas multiphase transportation system, especially on the multiphase flow patterns, multiphase pumping and multiphase metering of the system. A dynamic simulation analysis is conducted to deduce simulation parameters of the system and similarity criteria under simplified conditions are obtained. The reliability and feasibility of two-phase flow experiment with oil and natural gas simulated by water and air are discussed by using the similarity criteria.

Application of the Conjugate Gradient Optimization Algorithm of Neural Network on Laser Machining

在应用激光技术加工复杂曲面时,通常以采样点集为插值点来建立曲面函数,然后实现曲面上任意坐标点的精确定位。人工神经网络的BP算法能实现函数插值,但计算精度偏低,往往达不到插值精确要求,造成较大的加工误差。提出人工神经网络的共轭梯度最优化插值新算法,并通过实例仿真,证明了这种曲面精确定位方法的可行性,从而为激光加工的三维精确定位提供了一种良好解决方案。这种方法已经应用在实际中。

Numerical Simulation of Nox Formation in Coal Combustion with Inlet Natural Gas Burning

A full two-fluid model of reacting gas-particle flows and coal combustion is used to simulate coal combustion with and without inlet natural gas added in the inlet. The simulation results for the case without natural gas burning is in fair agreement with the experimental results reported in references. The simulation results of different natural gas adding positions indicate that the natural gas burning can form lean oxygen combustion enviroment at the combustor inlet region and the NOz concentration is reduced. The same result can be obtained from chemical equilibrium analysis.

Gas exchange by bubbles in waves

A simple and feasible model feet the calculation of the gas transfer by bubble clouds is proposed in this article. N-2, O-2, and CO2 transferred by bubble clouds are obtained. At wind speed of 10 m/s, the calculated supersaturation of dissolved oxygen is 1.93-3.89% in agreement with the measurement.

Flow properties of a dusty-gas point source in a supersonic free stream

By using Lagrangian method, the flow properties of a dusty-gas point source in a supersonic free stream were studied and the particle parameters in the near-symmetry-axis region were obtained. It is demonstrated that fairly inertial particles travel along oscillating and intersecting trajectories between the bow and termination shock waves. In this region,formation of "multi-layer structure" in particle distribution with alternating low- and highdensity layers is revealed. Moreover, sharp accumulation of particles occurs near the envelopes of particle trajectories.

Conjugate Model for Heat and Mass Transfer of Porous Wall in the High Temperature Gas Flow

Heat and mass transfer of a porous permeable wall in a high temperature gas dynamical flow is considered. Numerical simulation is conducted on the ground of the conjugate mathematical model which includes filtration and heat transfer equations in a porous body and boundary layer equations on its surface. Such an approach enables one to take into account complex interaction between heat and mass transfer in the gasdynamical flow and in the structure subjected to this flow. The main attention is given to the impact of the intraporous heat transfer intensity on the transpiration cooling efficiency.

Shock wave diffraction by a square cavity filled with dusty gas

A simple two-dimensional square cavity model is used to study shock attenuating effects of dust suspension in air. The GRP scheme for compressible flows was extended to simulate the fluid dynamics of dilute dust suspensions, employing the conventional two-phase approximation. A planar shock of constant intensity propagated in pure air over Aat ground and diffracted into a square cavity filled with a dusty quiescent suspension. Shock intensities were M-s = 1.30 and M-s = 2.032, dust loading ratios were alpha = 1 and alpha = 5, and particle diameters were d = 1, 10 and 50 mum. It was found that the diffraction patterns in the cavity were decisively attenuated by the dust suspension, particularly for the higher loading ratio. The particle size has a pronounced effect on the flow and wave pattern developed inside the cavity. Wall pressure historics were recorded for each of the three cavity walls, showing a clear attenuating effect of the dust suspension.