Modeling of fine coal flotation separation based on particle characteristics and hydrodynamic conditions

Flotation is a complex multifaceted process that is widely used for the separation of finely ground minerals. The theory of froth flotation is complex and is not completely understood. This fact has been brought many monitoring challenges in a coal processing plant. To solve those challenges, it is important to understand the effect of different parameters on the fine particle separation, and control flotation performance for a particular system. This study is going to indicate the effect of various parameters （particle Characteristics and hydrodynamic conditions） on coal flotation responses （flotation rate constant and recovery） by different modeling techniques. A comprehensive coal flotation database was prepared for the statistical and soft computing methods. Statistical factors were used for variable selections. Results were in a good agreement with recent theoretical flotation investigations. Computational models accurately can estimate flotation rate constant and coal recovery （correlation coefficient 0.85, and 0.99, respectively）. According to the results, it can be concluded that the soft computing models can overcome the complexity of process and be used as an expert system to control, and optimize parameters of coal flotation process.

Spectral transfer-learning-based metasurface design assisted by complex-valued deep neural network

Recently,deep learning has been used to establish the nonlinear and nonintuitive mapping between physical structures and electromagnetic responses of meta-atoms for higher computational efficiency.However,to obtain sufficiently accurate predictions,the conventional deep-learning-based method consumes excessive time to collect the data set,thus hindering its wide application in this interdisciplinary field.We introduce a spectral transfer-learning-based metasurface design method to achieve excellent performance on a small data set with only 1000 samples in the target waveband by utilizing open-source data from another spectral range.We demonstrate three transfer strategies and experimentally quantify their performance,among which the“frozen-none”robustly improves the prediction accuracy by∼26%compared to direct learning.We propose to use a complex-valued deep neural network during the training process to further improve the spectral predicting precision by∼30%compared to its real-valued counterparts.We design several typical teraherz metadevices by employing a hybrid inverse model consolidating this trained target network and a global optimization algorithm.The simulated results successfully validate the capability of our approach.Our work provides a universal methodology for efficient and accurate metasurface design in arbitrary wavebands,which will pave the way toward the automated and mass production of metasurfaces.

Amorphous particle deposition and product quality under different conditions in a spray dryer

Deposition of amorphous particles, as a prevalent problem particularly in the spray drying of fruit and vegetable juices, is due to low-molecular-weight sugars and is strongly dependent on the condition of the particles upon collision with the dryer wall. This paper investigates the condition of the amorphous particles impacting the wall at different drying conditions with the aim of elucidating the deposition mechanism and physical phenomena in the drying chamber. A model sucrose-maltodextrin solution was used to represent the low-molecular-weight sugar. Particle deposits were collected on sampling plates placed inside the dryer for analyses of moisture content, particle rigidity (using SEM) and size distribution. Moisture content was adopted as a general indicator of stickiness. Product particles collected at the bottom of the experimental dryer were found to have higher moisture than particle deposits on samplers inside the dryer. Moisture content profile in the dryer shows that apart from the atomizer region, where particles are relatively wet, particle deposits at other regions exhibit similar lower moisture content. At the highest temperature adopted in the experiments, particles became rubbery suggesting liquid-bridge formation as the dominant deposition mechanism. Further analysis on particles size distribution reveals a particle segregation mechanism whereby smaller particles follow preferentially to the central air stream while larger particles tend to re-circulate in the chamber, as predicted in past CFD simulation. The findings from this work will form the basis and provide validating data for further modeling of wall deposition of amorphous particles in spray drying using CFD.

Landmarks in the application of electrical tomography in particle science and technology

Selected milestones in the development and use of electrical tomography in powder conveying, slurry processing and multi-phase flow are highlighted. The ability to map concentration in opaque mixtures under process-realistic conditions was a major innovation for the method and has had far reaching implications. Subsequent developments have enabled velocity information to be abstracted resulting in the ability to measure component flux and motion.

Forecasting of dissolved oxygen in the Guanting reservoir using an optimized NGBM(1,1) model

An optimized nonlinear grey Bernoulli model was proposed by using a particle swarm optimization algorithm to solve the parameter optimization problem. In addition, each item in the first-order accumulated generating sequence was set in turn as an initial condition to determine which alternative would yield the highest forecasting accuracy. To test the forecasting performance, the optimized models with different initial conditions were then used to simulate dissolved oxygen concentrations in the Guantlng reservoir inlet and outlet （China）. The empirical results show that the optimized model can remarkably improve forecasting accuracy, and the particle swarm optimization technique is a good tool to solve parameter optimization problems. What＇s more, the optimized model with an initial condition that performs well in in-sample simulation may not do as well as in out-of-sample forecasting.