Research On The Identification Of Manual Adjustment Strategies And Intelligent Control Methods For Complex Processes

With the revolution in new energy technologies,the importance of nickel ore has become even more prominent.Whether it is solar,wind or other renewable energy technologies,nickel ore is required as a raw material.At the same time,the rapid development of the new energy vehicle market has intensified the demand for nickel ore.The shortage of nickel ore supply has become a constraint for new energy technologies and the new energy vehicle industry.As a result,the importance of ensuring the stability of nickel ore supply and increasing the amount of nickel ore available is becoming increasingly apparent.This has also become an important part of national and corporate development strategies.Froth flotation is a separation technique that takes advantage of differences in the hydrophilic properties of mineral surfaces.Hydrophilic mineral particles are deposited to the bottom of the flotation column and subsequently discharged to form tailings,while hydrophobic metallic minerals adhere to the air bubbles and form a stratification under buoyancy and collect at different locations in the flotation column,this difference separating and enriching the different components of the minerals.Today,froth flotation technology is widely used in the mining,chemical and metallurgical industries.In conventional flotation operations,there is too much reliance on the experience of the personnel on site,which is often susceptible to human factors and creates a high degree of subjectivity and uncertainty.This subjectivity and uncertainty can lead to large product quality fluctuations in the production process,which can affect the stability and consistency of production.In addition,due to the different experience levels and operating skills of personnel on site,different operators may have different interpretations and applications of dosing methods and dosing amounts,resulting in variability and uncertainty in the production process.A more scientific and intelligent control method is therefore needed to solve these problems.This paper relies on a collaborative project between the University of Electronic Science and Technology and Gansu Jinchuan Group,in which a large amount of data was generated.And some of this data is incomplete,inaccurate and under-utilised.In this project,a large amount of data from the mineral flotation database was analysed to filter the available data segments and process the invalid data segments to distinguish the different working conditions.For normal operating conditions,the adjustment of the main control parameters was observed,traces of adjustment were found and adjustment rules were generalised.In view of the low level of automation,the harsh environment and the subjective and arbitrary nature of worker control in the nickel froth flotation process,the experience of different workers is integrated and,based on advanced data analysis and model prediction techniques,the control scheme is automatically and intelligently optimised to output an effective adjustment strategy that simulates the manual adjustment process in mineral flotation.Ultimately,the relationships between control parameters,system parameters and target parameters of the froth flotation process are analysed based on the MLP multilayer perceptron-based discriminant model for the effectiveness of work parameters and adjustment strategies and the Transformer neural network-based prediction model for effective adjustment strategies,respectively.