Smart froth flotation
Many variables influence flotation performance, requiring constant monitoring and control of the process. Smart froth flotation solutions integrate advanced sensors, machine learning algorithms, and automation to deliver unprecedented insight and optimisation capabilities.
FLS’ froth camera with LiDAR is one such smart solution that represents the next generation of mineral processing vision systems. The cameras measure froth velocity and stability and can also infer froth grade using deep neural network (DNN) technology and object detection:
- DNN classifies the froth, such as barren, pulping, etc., while rejecting images affected by environmental conditions (such as direct sunlight) to eliminate false grade classifications.
- Object detection provides froth-grade detection by measuring froth velocity, bubble size, colour spectra, and mass pull rate.
LiDAR is an optional add-on to the froth camera that detects the froth height above and below the tip and creates a 3D representation of the froth zone. Froth height measurements are crucial when estimating mass and volumetric pull rates.
Froth cameras are complemented by PERI online slurry analysis, which uses a solid-state, room-temperature energy dispersive x-ray fluorescence (EDXRF) detector to provide accurate, reliable and cost-effective slurry analysis.
Data and insights from both froth cameras, online slurry analysis, and other instruments can then be fed into the FLS advanced process control (APC) solution, where they are combined with AI-based data analysis technologies, model predictive control, and fuzzy logic to optimise your flotation circuit proactively.
For example, FLS APC optimises the level and airflow targets using the mass pull rate distribution and camera signals. Mass pull rate control is based on the analyser values, downstream constraints and other plant targets. This allows you to perfectly stabilise flotation cell level control, reducing variability in the cell and maximising recovery at the desired grade.
