S. Joe Qin – Inferential Monitoring of Quality and Process
Faults with Supervised Learning

Inferential Monitoring of Quality and Process Faults with Supervised Learning

Principal component analysis and unsupervised machine learning models have been applied to process monitoring to analyze variations in process variables and capture the normal region of variability. In practice, however, anomalies detected in process variables alone do not always result in an anomaly in product quality due to corrective actions by human operators and feedback controllers. Concurrent monitoring schemes have been actively studied recently to achieve simultaneous process and quality-relevant monitoring.

In this talk we introduce a new framework of inferential monitoring and diagnosis (IMD) to include quality-relevant monitoring and quality-irrelevant monitoring schemes using supervised learning models such as partial least squares, canonical correlation analysis (CCA), and other machine learning methods. It is related to and complementary to the scheme of inferential sensors or soft sensors. Monitoring indices based on supervised learning models are given to perform inferential monitoring.

Additionally, contribution plots, reconstruction-based contributions, and generalized reconstruction-based contribution methods are adopted for inferential diagnosis. Finally, industrial case studies and the Tennessee Eastman process are used to illustrate the inferential monitoring and diagnosis of quality-relevant and quality-irrelevant faults.




Prf. Chris Aldrich – Recent Advances in Multivariate Image Analysis
in the Mineral Processing Industries

Recent Advances in Multivariate Image Analysis
in the Mineral Processing Industries

Recent developments in deep learning has generated a new impetus in the development of multivariate image analytical methods that are applied in the development of smart sensors, process monitoring and advanced control of process systems. This includes advances in froth image analysis, ore sorting, monitoring of particulate feeds on conveyor belts, state identification of hydrocyclone operations, but are also opening up potentially new avenues in the development of autonomous inspection systems.

Moreover, these approaches are not confined to the processing of optical images, but can also be applied to signal processing in general. In this overview, the latest developments, challenges and future trends in these methods will be considered, with an emphasis on the use of deep learning and emerging applications in the mineral processing industries.








O.A. Bascur * and A. Soudek – Process Analytics: Transforming Mineral Process Plant Data Into Actionable Insights

Process Analytics: Transforming Mineral Process Plant Data Into Actionable Insights

The advent of digital revolution has now enabled us with numerous tools that could be leveraged to transform our operational data into actionable insights.
Key opportunities with digitization include better visualization, transparency, integrated planning and execution for value-chain optimization, that results in smarter production, intelligent response to changes in ore, process and equipment conditions, reduce energy and waste along with prevention of asset breakdown, safety and environmental issues.

It is important to realize that these digital tools have limited value, from a metallurgical operational context, if we cannot bring-in the appropriate domain expertise along with getting the basics right. The focus is to ensure that there is adequate depth of different disciplines built into our platforms along with breadth to integrate other disciplines such as geology and mining for an effective Mine-to-Mill integration.

A metal recovery improvement strategy based on optimal Gaudin particle size distribution, air hold up manipulation, integrated grinding-flotation performance monitoring and guidance is discussed. Simultaneous identification of process and equipment constraints enable finding the best overall conditions for metal recovery. Several industrial success stories are presented.

This paper discusses the methodology, findings, and challenges in the ongoing journey of implementing a Metallurgy Analytics platform that evolves from being retrospectively descriptive to anticipatively prescriptive.


Kathryn Hadler – Flotation Flowsheet Design And Optimisation

Flotation Flowsheet Design And Optimisation 

Flowsheet design is the most creative part of process engineering. The goal of the designer is to develop a flowsheet that produces the maximum amount of a consistent, specific quality product from a given feedstock.  This requires a wide range of options to be considered, evaluated and the most appropriate selected, under various constraints.

Plant designers and equipment manufacturers minimise risk, with the result that comminution and flotation plant flowsheets have remained largely unchanged, except that equipment has grown in size.  Unfortunately, flotation tank size has now reached the upper limit with respect to kinetic constraints.  The question remains whether the “standard” rougher-scavenger-cleaner circuit is the optimal flowsheet layout.

Flowsheet design can be optimised using genetic algorithms.  This requires an appropriate flotation model that includes particle size, liberation and kinetics, and froth properties and air rate effects.  The development of a flotation circuit optimisation system using genetic algorithms will be described in this presentation. The results show that the number of possible layouts depends exponentially on the number of tanks, and that without appropriate constraints the combinatorial problem can become intractable.

Such constraints will be described. It is shown that the design system is able to produce realistic and practical layouts, where the position of recycle streams into the rougher bank is determined optimally.

Of particular interest is the finding that the optimal circuit layout depends on the feed particle size, liberation and floatability model. The results show that circuits tend towards the familiar flotation circuits as feed model complexity increases. The implication of this with regards to modelling of flotation and to flotation operation will be discussed, in addition to a forward look at the design of mineral processing circuits.


Prof. Jan Cilliers – Flotation optimisation using Peak Air Recovery: Past, present and future

Flotation optimisation using Peak Air Recovery: Past, present and future

Froth properties determine flotation performance. The optimisation, automation and control of froth behaviour and flotation performance remains one of the great challenges in mineral processing today. One solution to this is Peak Air Recovery (PAR), a simple concept in which a single target (air recovery) is maximised by varying physical operating parameters such as air rate.

Around 20 years ago, fundamental models of flowing, coalescing froths and the distribution of liquid, gas and solids identified an essential mathematical boundary condition; the fraction of air entering the flotation cell that overflows as unburst bubbles. This concept, called air recovery, accounts both for bubble stability and froth mobility. It can be measured readily on plant, as it is the ratio of volume rate of froth overflowing (froth velocity x overflowing froth height above the lip x lip length) to the feed gas volume rate (superficial gas velocity x cell area). Although the importance of air recovery was discovered in early froth physics models, its real value has been demonstrated through numerous plant surveys carried out over the last 15 years. The values of air recovery measured on plant are surprisingly low; generally less than 50%, and often much lower. More significantly, we have shown that mineral recovery can be maximised, often at higher concentrate grade, when operating flotation cells and circuits at their optimal air recovery.

The maximum, Peak Air Recovery (PAR), therefore gives a single target for a flotation control strategy. This has been trialled industrially by Rio Tinto and Anglo Platinum.

In this presentation, the concept of PAR, its variability and plant trial results will be reviewed. The development of a laboratory-based control system will be described. Furthermore, drawbacks to the implementation of an air recovery-based control system will be discussed, including plant instability and lack of instrumentation. On the latter, a simplified method that avoids the need for accurate inlet air rate – often a limiting factor for many plants – will be presented. Finally, the outlook for air recovery as a control strategy will be addressed.