Industrial Flotation Plant Research Articles

On-Line Image Analysis to Improve Industrial Flotation Plant Performance

In this study image features extracted from froth images by an on-line machine vision system in an industrial platinum flotation plant were used to relate froth characteristics and flotation performance by using neural networks. It has been shown that a considerable amount of data can be extracted from flotation surface froths and that both novel feedback control procedures and feedback control as a complement to conventional feedforward systems are made possible. Feature measures such as chromatic information, average bubble size, froth texture, froth stability and mobility of surface froths were used in the on-line classification of flotation froths. This intelligent vision system constitutes a powerful research tool for the investigation and interpretation of the effect of various flotation parameters. This paper shows how the rapid development in computer technology and neural networks can be used to transform recently developed concepts and available technology into a new generation of intelligent automation systems.

The monitoring of froth surfaces on industrial flotation plants using connectionist image processing techniques

The rapid development of computer vision, computational resources, artificial intelligence and the integration of these technologies are creating new possibilities in the design and implementation of commercial machine vision systems. In minerals engineering numerous opportunities for the application of these systems exist, such as the characterization of flotation froth structures which is discussed in this paper by way of example. A general model for the development of feasible, real-time machine vision systems is proposed, which is based on an analogy with biological visual perception made possible by a connectionist approach and the ability of neural networks to solve ill-posed problems. It is shown that both supervised and unsupervised neural nets can be used in different ways to analyze froth structures of industrial flotation cells. Unsupervised (self-organizing) neural nets can monitor process behaviour on a continuous rather than on a discrete basis, which makes the early detection of erratic process control possible. Since some losses in information are incurred with the use of self-organizing systems, intelligent monitoring and control systems would in practice probably be comprised of both types of neural nets.

Digital image processing as a tool for on-line monitoring of froth in flotation plants

As the most important separation technique in mineral processing, flotation has been the subject of intensive investigation over many years, but despite these efforts it remains a poorly understood process that defies generally useful mathematical modelling. As a result the control of industrial flotation plants is often based on the visual appearance of the froth phase, and depends to a large extent on the experience and ability of a human operator. These types of processes are consequently often controlled suboptimally owing to high personnel turnover, lack of fundamental understanding of plant dynamics, inaccuracy or unreliability of manual control systems, etc. By using techniques based on image colour analysis and Fast Fourier Transforms to process videographic data of the froth phase in a copper flotation plant, it is shown that an image processing system can distinguish between different copper levels in the froth down a rougher bank and extract global features from the visual characteristics of the surface froth. In this way it is possible to quantify the mineral content of the froth (based on colour), the average bubble size distribution, the direction of flow and the shape of the bubbles or the mobility of the froth. The overall image of the froth is analysed instead of attempting to identify the boundaries between bubbles.

Identification of feed and simulation of industrial flotation circuits

Feed to flotation circuits is characterized by conjugate distributions of total mass and mineral mass in apparent flotation rate constant as the common attribute. The rate constant is interpreted as a lumped kinetic parameter in a fairly detailed multi-phase flotation model which explicitly incorporates bubble phase and froth stability index. A method is described for identification of the discretized conjugate distributions from flotation plant data. The procedure entails random search of optimum feed parameters in conjunction with repeated simulation of the circuit in two parallel mathematical streams, one for total mass flows and second for mineral flows only. Feeds to three industrial flotation plants have been identified for which flotation data are available in literature. The plants could be simulated stage-wise quite accurately with two to five species description of the feed. From these results and extensive trials based on computer-generated flotation data, it is concluded that the inverse problem of feed identification can be solved. The identified feed is meaningful for circuit simulation, provided the flotation rate is scaled in each stage of the circuit to account for variations in pulp density and chemistry, residence time, aeration rate and froth stability. Some applications are discussed.

The use of linear programming in the optimal design of flotation circuits incorporating regrind mills

Two linear models, the second being a subset of the first, are proposed for the simulation of flotation plants by use of linear programming. The first linear model produces the circuit structure, as well as the optimal flow rates of the valuable element between any number of flotation banks incorporating any number of recycle mills. An optimal grade for the valuable element in the concentrate is given by the second model. Operating conditions in the flotation banks and recycle mills are included as bounds in these models, permitting their possible application in expert systems. The simulated circuit structure, concentrate grade and recoveries closely resemble those of similar industrial flotation plants. The only data required by the simulation models are the feed rates of the species of an element, and their separation factors which are estimated from a multiparameter flotation model.