Model-based soft sensors for comminution process control and optimisation

Speaker

Tania Ledezma

Abstract

This seminar will present the development and application of model-based soft sensors as advanced tools for monitoring and controlling comminution circuits. Soft sensors are virtual instruments that infer key process variables using mathematical models and real-time data, offering a robust alternative to traditional hardware-based instrumentation, particularly in scenarios where direct measurement is difficult or impractical.
The presentation will introduce the concept of soft sensors and discuss the distinctions between data-driven, model-based, and hybrid approaches. Emphasis will be placed on the methodology developed at the Julius Kruttschnitt Mineral Research Centre (JKMRC), which relies on validated physical models to ensure reliability, interpretability, and adaptability across diverse operational contexts. The talk will highlight several case studies demonstrating the implementation of tools such as JK MillFIT, JK CycloPS, and JK VMillFIT, which have provided accurate, real-time estimates of critical variables including mill filling, cyclone performance, and particle size distribution. These tools have been successfully integrated into modern digital environments, enabling enhanced process control, predictive analytics, and operational optimisation.

Bio

Tania Ledezma is a researcher at the Julius Kruttschnitt Mineral Research Centre (JKMRC) at the Sustainable Minerals Institute, The University of Queensland. Her areas of expertise are process control, modelling, and soft sensor development for mineral processing. With over nine years of experience in the Chilean mining industry, she has held roles in process engineering, data analysis, and control systems. Tania holds a master’s degree in chemical engineering with a speciality in process control and has contributed to both industrial innovation and academic collaboration. Her work supports the integration of digital technologies to enhance operational efficiency and process understanding in the mining sector.