Industrial Froth Research Articles

Froth velocity extraction method from froth image for flotation with large throughput based on composite deep learning network

Froth velocity is a crucial variable in froth flotation and its accurate detection is vital to process control. The velocity extraction is a challenging work under large throughput conditions due to the complex motion of the froth. Therefore, in this paper, a novel method called composite deep learning network (CDLN) is proposed, which is mainly composed by a cascade of pre-trained supervised network and an unsupervised network. The supervised network is pre-trained on a benchmark dataset and then is used to extract an initial flow field of the froth motion. The unsupervised network is trained on industrial froth videos with occlusion modeling to get a calibration flow field. The two flow fields are then fused to obtain the final flow field. The test results show that the proposed method can extract the froth velocity more accurately and faster than other popular methods through industrial froth videos under large throughput.

Working condition perception for froth flotation based on NSCT multiscale features

Visual features at the froth surface of a flotation cell are closely related to the process condition and performance. Accurate identification of flotation conditions is crucial. However, the multi-scale properties of froth images are not all extracted based on existing perception methods. Considering that multi-scale frequency domain analysis can obtain more informative statistical froth features, a new working condition recognition based on nonsubsampled contourlet transform (NSCT) multiscale features is proposed and successfully validated by industrial froth images taken under different process conditions. NSCT was used to enhance the froth image, including the decomposition and reconstruction, contrast-limited adaptive histogram equalization (CLAHE), and multi-scale features of the enhanced image are extracted and analyzed. The fuzzy binarization method is adopted for extracting the morphological feature. Texture feature parameters are extracted from other high-frequency images. The particle swarm optimization support vector machine is used to recognize the working condition. The results indicate that the proposed algorithms, in particular the multiscale features classification based on NSCT, have the highest classification accuracy (93.33 %), and can accurately and reliably identify the working condition in the actual froth images.

Deep architecture for silica forecasting of a real industrial froth flotation process

Froth Flotation is one of the fundamental processes used in iron ore attainment. It involves various physical–chemical reactions, and its multivariate control is not trivial. Traditionally, plant operators get quality measurements from laboratory analysis of sampled material, which can take hours to be published. In this paper, we present alternatives that eliminate both the necessity to periodically sample material as well as the delay associated with laboratory analysis: a predictive machine learning model working as a soft sensor. Two different classes of models were developed, exploring both deep and shallow approaches of machine learning, with a deep neural network architecture constructed for the particular task at hand. All models yielded accurate predictions, with results favoring the deep neural network when a larger volume of data is available.

Feature Reconstruction-Regression Network: A Light-Weight Deep Neural Network for Performance Monitoring in the Froth Flotation

With the rapid development of deep neural network (DNN), many DNN-based models for performance monitoring have been developed recently. However, some challenges still exist in the industrial performance monitoring: 1) different sample rates and time delays between the inputs and labeled performance; 2) a light-weight DNN architecture. Under this circumstance, we design a DNN named feature reconstruction-regression network (FR-R net) in this article. First, we extract the feature vector series as the input feature in order to capture the dynamic temporal information of the input data. Then, we design a feature reconstruction network with a weight-shared kernel network and fixed positional encoding to generate a reconstructed feature vector. Finally, we send the reconstructed feature vector into fully connected layers as a regression network to link the labeled performance. The effectiveness of the proposed FR-R net is validated on both a simulation case and an industrial froth flotation process.

Experimental observations of bubble–particle collisional interaction relevant to froth flotation, and calculation of the associated forces

This paper modifies an experimental system previously used by Verrelli et al. to investigate particle–bubble attachment. Three different flow configuration cases were studied to investigate the effects of flow environment on the particle–bubble interactions and to model a situation more characteristic of real industrial froth flotation processes. The three configurations are as follows: particles dropped onto a bubble held stationary at the tip of a capillary tube; particles in water flowing past a bubble held stationary; and particles dropped onto a bubble rising freely. This is the first time collisional interaction between a freely rising bubble and hydrophobized glass ballotini particles has been studied in detail. The present work has shown the importance of considering real flow situations when modelling bubble–particle interactions for flotation applications. According to the results, there are more particles swept off the bubble surface before reaching the 90° position from the vertical axis under conditions of liquid flow (either because of externally generated flow or because of bubble rise), compared to the case where the bubble is stationary in a stationary liquid. This appears to suggest that attachment probability is reduced when liquid flows past the bubble. Extracting values of the individual force components provides greater insight into the mechanisms influencing the particle’s behavior as it interacts with the bubble. Comparison of particle trajectories that resulted in attachment with those that did not, provides further evidence in support of the proposition that contact time needs to be longer than induction time for attachment to occur.

A Computational and Experimental Study on the Binding of Dithio Ligands to Sperrylite, Pentlandite, and Platinum

This benchmark study documents a combined computational and experimental investigation into the binding of three commercial dithio collector ligands used in industrial froth flotation processes to separate high-value minerals from lower-value materials. First-principles condensed-matter simulations showed that ethyl xanthate, N,N-diethyl dithiocarbamate, and diisobutyl dithiophospinate anions all bind least strongly to the [100] Miller index plane of the platinum-containing mineral sperrylite, followed by the [111] MI surface of the mixed nickel/iron sulfide mineral pentlandite, whereas all ligands showed the strongest binding affinity to the [111] MI surface model of pure platinum. Calculations also support experimental observations that neutral ethyl xanthogen disulfide formed upon oxidation of ethyl xanthate binds much more weakly than the monomer. A monolayer of water molecules would easily be displaced from all surfaces by any of the collector ligands. The hydroxide anion was found to have binding en...

Optical luminescence studies of diffusion times at the potassium ethyl xanthate adsorption layer on the surface of sphalerite minerals

We propose reflectance measurements as a method for the evaluation of the kinetics of adsorption processes, to compute the diffusion times of the adsorption products at the thin layers formed at the sphalerite natural mineral-potassium ethyl xanthate solution interface. The method is based on the intensity measurement of the reflected monochromatic radiation obtained from the mineral-xanthate thin layer as a function of time. These determinations were made at the thin layer formed between the sphalerite or activated sphalerite natural minerals with potassium ethyl xanthate, for different solutions concentrations and pH values at constant temperature. Diffusion times of desorbed molecular species into the liquid bring important information about the global kinetics of the ions in this phase during adsorption processes at interfaces. Analysing the time dependence of this parameter one concluded on the diffusion properties of the xanthate molecule in the solution depending on its concentration and pH, knowing that at the initial time these molecules had a uniform spread. This method enabled us to determine that, in time interval of approximately 35 minutes to achieve dynamic equilibrium in the formation of the interface layer, one had three different kinetic behaviours of our systems. In the first 5-8 min one had highly adsorbent character, the state of equilibrium is followed by low adsorbent properties. Gaining information on the adsorption kinetics in the case of xanthate on mineral surface leads to the optimization of the industrial froth flotation process.

Optical luminescence studies of the ethyl xanthate adsorption layer on the surface of sphalerite minerals

In this work we propose optical luminescence measurements as a method to evaluate the kinetics of adsorption processes. Measurement of the intensity of the integral optical radiation obtained from the mineral–xanthate interface layer, stimulated with a monochromatic pulsating optical signal, as a function of time were made. The luminescence radiation was obtained from the thin interface layer formed at the separation surface between the sphalerite natural mineral and potassium ethyl xanthate solution, for different solution concentrations and pH-es at the constant industry standard temperature. This method enabled us to determine the time to achieve dynamic equilibrium in the formation of the interface layer of approximately 20min, gaining information on the adsorption kinetics in the case of xanthate on mineral surface and leading to the optimization of the industrial froth flotation process.

Flotation process fault detection using output PDF of bubble size distribution

As an important indicator of flotation performance, bubble structure is believed to be related with the addition amount of chemical reagent. To differentiate film size calculated from surface froth image segmentation and underlying bubble size, post-segmentation analysis is carried out to archive the prediction of bubble size distribution from the known bubble film distribution. Based on the facts that the probability density distribution (PDF) of film size is found to be non-Gaussian and singular feature characterizing bubble structure becomes insufficient, a nonparametric kernel estimator is introduced to approximate the output PDF of bubble size distribution. Through transforming nonparametric description into dynamic kernel weight vectors, an output PDF based detection filter is designed, according to which LMIs are established to solve its stability condition by Lyapunov stability analysis on the estimated error system. Fault detection problem is then solved through a threshold criterion determined by the system stability condition. Simulation results show that desired fault detection for reagent addition in industrial froth flotation can be achieved using the proposed method.

Interpretation of the effect of frother type and concentration on flotation performance in an OK3 cell

Over the past few decades a great deal of effort has gone into characterizing the effects of frother type and concentration in flotation using a variety of two-phase techniques. Whilst useful and informative, this research still does not assist in the selection of suitable frothers for three-phase industrial-scale froths. This paper investigates the effects of frother type and concentration in a three-phase industrial froth. The objective was to determine if progress could be made in terms of understanding the behavior of frothers in flotation and understanding how they affect the sub-processes occurring in the pulp and froth phases and, ultimately, the metallurgical performance of a flotation cell.

An image processing algorithm for measurement of flotation froth bubble size and shape distributions

The appearance of the upper bubble surface of a flotation froth is used qualitatively as a basis for control in industrial flotation cells. The complex structures that constitute the surface of a froth are indicative of the sub-processes occurring within the froth, and which affect the flotation performance. Being able to accurately and automatically measure bubble size and shape distributions is therefore an important requirement for the optimisation and control of froth processes. An algorithm has been developed for froth surface bubble size distribution measurement, and that is largely insensitive to factors such as froth type, lighting conditions and bubble size. In this algorithm, reconstruction is applied following the subtraction of a constant value from the image. Regional maxima are identified and used as markers for the watershed, which segments the image. Equivalent circular bubble size and shape distributions can be determined from the number of pixels in each segment. It is shown that the bubble segmentation algorithm can be generally applied to flotation froths from a variety of minerals and conditions, and may be suitable for application in automated industrial froth characterisation and control systems.