Intelligent Control Study Of Flotation Based On Optimization Model Of Dosing Parameters

As one of the processes with complex operation mechanism and many influencing variables in the field of coal beneficiation,flotation brings difficulties in analyzing flotation process and optimizing flotation process control parameters to improve flotation process efficiency due to its inherent characteristics of nonlinearity and strong coupling,as well as the limitation of certain process parameter detection technology problems.In this thesis,we establish a flotation process parameter detection platform,build a flotation dosing control parameter optimization model,and introduce a feedback compensation control strategy to investigate the establishment of an accurate and stable flotation dosing feedback compensation intelligent control system.A study of flotation influencing variables and process data collection methods was conducted for a coal processing plant.Firstly,by analyzing the variables of flotation process,the raw coal quantity,coal type,feed ash,feed flow rate,feed concentration,chemical addition,as well as flotation concentrate ash and tailing ash were selected as variables for analyzing the flotation process.After that,the data detection methods of each variable were analyzed,and a flotation process parameter detection platform was built using C# language and SQL database.The data set collected by the platform is processed for time dimension unification,outliers and missing values.Firstly,the time dimension is unified with the whole point moment of floating fine gray sampling,after that,the outliers are removed by combining the box line graph method and the determination of threshold range method,and the missing values are filled and removed to some extent by using the EM algorithm and the direct rejection method.The effective utilization rate of the final data sample reaches 85.25%,which proves the effectiveness of the data processing method.A study of a non-analytic prediction model for flotation ash was initiated.A neural network-based non-analytical prediction model for flotation ash was proposed to address the problem of difficulty in establishing an accurate mathematical model to describe the flotation input-output relationship.Firstly,the model input and output variables were determined,and the data were standardized using the polar difference standardization method.After that,the flotation ash non-analytic prediction models based on BP and Elman neural networks were developed respectively,and the structure of both models was "7 inputs,2 outputs".The results show that the Elman-based prediction model has obvious advantages in terms of prediction accuracy and stability for flotation ash and tailing ash.In order to achieve accurate control of the flotation ash content of the system,a study on the optimization model of flotation dosing parameters was carried out.Firstly,the actual production status was analyzed,and two control parameters of frothing agent and trapping agent addition were selected as optimization variables.After that,based on the Elman ash prediction model,a genetic algorithm was used to construct the model,and the model process was as follows: iteratively updating the agent addition instead of the original data,forming new samples to input into the ash prediction model,and finally obtaining the optimized agent addition.The root mean square error was 0.877 and 0.410 for the two optimization methods,which showed that the second optimization method was more optimized and the optimized dosage had practical significance,which proved the effectiveness of the optimization model.Based on the feedback compensation control strategy,a set of flotation dosing feedback compensation intelligent control system is established.Firstly,the feedback compensation controller is designed by introducing fuzzy control for the optimization model and the errors of chemical addition caused by various equipments in practice.After that,the system architecture of the control system is designed and the system program design of the corresponding upper computer and lower computer is completed to complete the construction of the flotation dosing feedback compensation intelligent control system.