Modeling phytoplankton dynamics in the River Darling (Australia) using the radial basis function neural network

A radial basis function neural network was employed to model the abundance of cyanobacteria. The trained network could predict the populations of two bloom forming algal taxa with high accuracy, Nostocales spp. and Anabaena spp., in the River Darling, Australia. To elucidate the population dynamics for both Nostocales spp. and Anabaena spp., sensitivity analysis was performed with the following results. Total Kjeldahl nitrogen had a very strong influence on the abundance of the two algal taxa, electrical conductivity had a very strong negative relationship with the population of the two algal species, and flow was identified as one dominant factor influencing algal blooms after a scatter plot revealed that high flow could significantly reduce the algal biomass for both Nostocales spp. and Anabaena spp. Other variables such as turbidity, color, and pH were less important in determining the abundance and succession of the algal blooms.

Modeling and optimization of semiconductor manufacturing process with neural networks

A neural network-based process model is proposed to optimize the semiconductor manufacturing process. Being different from some works in several research groups which developed neural network-based models to predict process quality with a set of process variables of only single manufacturing step, we applied this model to wafer fabrication parameters control and wafer lot yield optimization. The original data are collected from a wafer fabrication line, including technological parameters and wafer test results. The wafer lot yield is taken as the optimization target. Learning from historical technological records and wafer test results, the model can predict the wafer yield. To eliminate the "bad" or noisy samples from the sample set, an experimental method was used to determine the number of hidden units so that both good learning ability and prediction capability can be obtained.

Retention modeling and simultaneous optimization of pH value and gradient steepness in RP-HPLC using feed-forward neural networks

A novel approach is proposed for the simultaneous optimization of mobile phase pH and gradient steepness in RP-HPLC using artificial neural networks. By presetting the initial and final concentration of the organic solvent, a limited number of experiments with different gradient time and pH value of mobile phase are arranged in the two-dimensional space of mobile phase parameters. The retention behavior of each solute is modeled using an individual artificial neural network. An "early stopping" strategy is adopted to ensure the predicting capability of neural networks. The trained neural networks can be used to predict the retention time of solutes under arbitrary mobile phase conditions in the optimization region. Finally, the optimal separation conditions can be found according to a global resolution function. The effectiveness of this method is validated by optimization of separation conditions for amino acids derivatised by a new fluorescent reagent.

Retention modeling and optimization of pH value and solvent composition in HPLC using back-propagation neural networks and uniform design

A novel method for the optimization of pH value and composition of mobile phase in HPLC using artificial neural networks and uniform design is proposed. As the first step. seven initial experiments were arranged and run according to uniform design. Then the retention behavior of the solutes is modeled using back-propagation neural networks. A trial method is used to ensure the predicting capability of neural networks. Finally, the optimal separation conditions can be found according to a global resolution function. The effectiveness of this method is validated by optimization of separation conditions for both basic and acidic samples.

Using the moving window incorporated neural network to forecast the population behavior of Nostocales spp. in the River Darling, Australia

The paper demonstrates the nonstationarity of algal population behaviors by analyzing the historical populations of Nostocales spp. in the River Darling, Australia. Freshwater ecosystems are more likely to be nonstationary, instead of stationary. Nonstationarity implies that only the near past behaviors could forecast the near future for the system. However, nonstionarity was not considered seriously in previous research efforts for modeling and predicting algal population behaviors. Therefore the moving window technique was incorporated with radial basis function neural network (RBFNN) approach to deal with nonstationarity when modeling and forecasting the population behaviors of Nostocales spp. in the River Darling. The results showed that the RBFNN model could predict the timing and magnitude of algal blooms of Nostocales spp. with high accuracy. Moreover, a combined model based on individual RBFNN models was implemented, which showed superiority over the individual RBFNN models. Hence, the combined model was recommended for the modeling and forecasting of the phytoplankton populations, especially for the forecasting.

Modeling of cyanobacterial blooms in hypereutrophic Lake Dianchi, China

Compared with other approaches for modeling and predicting, artificial neural networks are more effective in describing complex and non-linear systems. The occurrence of cyanobacterial blooms has been a continuous and serious problem over the past decades in hypereutrophic Lake Dianchi. Yet, the main factor(s) initiating these blooms remain(s) unclear. During 2001-2002 at 40 sampling sites in Lake Dianchi, physicochemical parameters possibly relating to the blooms were measured. Parameters directly or indirectly relating to the cyanobacterial blooms were used as driving factors in a back-propagation network to model the concentration of chlorophyll a. According to sensitivity analysis, chemical oxygen demand was identified as a very significant environmental factor for algal growth in Lake Dianchi.

The image coding scheme based on neural network and SVM

This paper applies data coding thought, which based on the virtual information source modeling put forward by the author, to propose the image coding (compression) scheme based on neural network and SVM. This scheme is composed by "the image coding (compression) scheme based oil SVM" embedded "the lossless data compression scheme based oil neural network". The experiments show that the scheme has high compression ratio under the slightly damages condition, partly solve the contradiction which 'high fidelity' and 'high compression ratio' cannot unify in image coding system.

Modeling and mesoscopic damage constitutive relation of brittle short-fiber-reinforced composites

Aimed at brittle composites reinforced by randomly distributed short-fibers with a relatively large aspect ratio, stiffness modulus and strength, a mesoscopic material model was proposed. Based on the statistical description, damage mechanisms, damage-induced anisotropy, damage rate effect and stress redistribution, the constitutive relation were derived. By taking glass fiber reinforced polypropylene polymers as an example, the effect of initial orientation distribution of fibers, damage-induced anisotropy, and damage-rate effect on macro-behaviors of composites were quantitatively analyzed. The theoretical predictions compared favorably with the experimental results.

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.

Modeling Study on the Characteristics of Laminar and Turbulent Argon Plasma Jets Impinging Normally upon a Flat Plate in Ambient Air

Modeling study is performed to compare the flow and heat transfer characteristics of laminar and turbulent argon thermal-plasma jets impinging normally upon a flat plate in ambient air. The combined-diffusion-coefficient method and the turbulence-enhanced combined-diffusion-coefficient method are employed to treat the diffusion of argon in the argon-air mixture for the laminar and the turbulent cases, respectively. Modeling results presented include the flow, temperature and argon concentration fields, the air mass flow-rates entrained into the impinging plasma jets, and the distributions of the heat flux density on the plate surface. It is found that the formation of a radial wall jet on the plate surface appreciably enhances the mass flow rate of the ambient air entrained into the laminar or turbulent plasma impinging-jet. When the plate standoff distance is comparatively small, there exists a significant difference between the laminar and turbulent plasma impinging-jets in their flow fields due to the occurrence of a large closed recirculation vortex in the turbulent plasma impinging-jet, and no appreciable difference is found between the two types of jets in their maximum values and distributions of the heat flux density at the plate surface. At larger plate standoff distances, the effect of the plate on the jet flow fields only appears in the region near the plate, and the axial decaying-rates of the plasma temperature, axial velocity and argon mass fraction along the axis of the laminar plasma impinging-jet become appreciably less than their turbulent counterparts.

Effect of Rock Mass Structure and Block Size on the Slope Stability-Physical Modeling and Discrete Element Simulation

This paper studies the stability of jointed rock slopes by using our improved three-dimensional discrete element methods (DEM) and physical modeling. Results show that the DEM can simulate all failure modes of rock slopes with different joint configurations. The stress in each rock block is not homogeneous and blocks rotate in failure development. Failure modes depend on the configuration of joints. Toppling failure is observed for the slope with straight joints and sliding failure is observed for the slope with staged joints. The DEM results are also compared with those of limit equilibrium method (LEM). Without considering the joints in rock masses, the LEM predicts much higher factor of safety than physical modeling and DEM. The failure mode and factor of safety predicted by the DEM are in good agreement with laboratory tests for any jointed rock slope.

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。

Optical Diagnostic and Modeling Solution Growth Process of Sodium Chlorate Crystals

Both a real time optical interferometric experiment and a numerical simulation of two-dimension non-steady state model were employed to study the growth process of aqueous sodium chlorate crystals. The parameters such as solution concentration distribution, crystal dimensions, growth rate and velocity field were obtained by both experiment and numerical simulation. The influence of earth gravity during crystal growth process was analyzed. A reasonable theory model corresponding to the present experiment is advanced. The thickness of concentration boundary layer was investigated especially. The results from the experiment and numerical simulation match well.

A Dynamic Loading Device for Suction Foundations in Centrifuge Modeling

Suction bucket foundations are widely used in the offshore platform for the exploitation of the offshore petroleum and natural gas resources. During winter seasons, ice sheets formed in Bohai Bay will impose strong impact and result in strong vibration on the platform. This paper describes a dynamic loading device developed on the geotechnical centrifuge and its application in modeling suction bucket foundation under the equivalent ice-induced vibration loadings. Some experimental results are presented. It is shown that when the loading amplitude is over a critical value, the sand at the upper part around the bucket softens or even liquefies. The excess pore pressure decreases from the upper part to the lower part of the sand foundation in vertical direction while decreases from near to far away from the bucket's side wall in the horizontal direction. Large settlements of the bucket and the sand around the bucket occur under the horizontal dynamic loading. The dynamic responses of the bucket with smaller size are heavier.