Study and application of autonomous adjustment system for flotation process control based on multivariable coupling technology

With an aim to tackle the problems faced by Tianchen Coal Mine's coal preparation plant in operation of its flotation process control — including limited diversity in agent adjustment, unstable indicators, agent waste and high manual labor intensity — an autonomous adjustment system based on multivariable coupling technology is developed. Based on the production status of the flotation system at Tianchen coal preparation plant, the multivariable coupling relationships in coal slime flotation are analyzed, clarifying the interactions among factors such as feed characteristics and agent regime. An integrated measuring device for pulp level and froth layer limit is developed to achieve real-time collection and monitoring of pulp level and froth thickness. The automation upgrade of the pulp level adjustment mechanism is carried out to stabilize pulp level control in flotation cells. An aeration rate measurement and control device is added to enable online regulation of aeration volume. Multiple systems are integrated to establish an autonomous adjustment system for flotation process control, with a B/S architecture-based software platform developed for centralized monitoring and data interoperability. A product index prediction model is constructed to forecast tailings ash content, achieving a mean square error of 1.56% in tailings ash prediction. Application results demonstrate that the overall flotation performance improved significantly after implementing the autonomous adjustment system for flotation process control. The clean coal yield increased from 76.36% to 77.42%, generating an annual economic benefit of 1.285 million yuan. Through the implementation of this adjustment system, Tianchen coal preparation plant effectively optimizes flotation performance. Despite deteriorating feed coal quality, the system enhances product stability, reduces reagent consumption, and increases clean coal yield, delivering tangible economic, social, and environmental benefits. With future improvements in ash data reliability, the predictive model precision is projected to advance continuously, providing robust support for intelligent flotation process upgrades.