Research On Operation Situation Awareness Method Of Mineral Flotation Process Based On Multi-source Data

Due to the fluctuation of mineral sources and the change of relevant operating parameters,the flotation working condition often changes frequently during the mineral flotation process.To make the flotation process run stably,it is an important prerequisite to accurately perceive the operation state and change the trend of the flotation process.Situation awareness can obtain the changing trend of system operation in advance and avoid system fluctuation.Therefore,combining the situation awareness,this paper designs a situation awareness scheme for the flotation process,that has great significance for comprehensively perceiving operation situations and changing trends of the flotation process and for maintaining a stable and efficient flotation production process.The main research contents and innovations are as follows:(1)A flotation condition evaluation method based on cloud modelimproved evidence theory is proposed to solve the problem that the uncertainty of multi-source element data makes it difficult to accurately evaluate the operation situation of the flotation process.The membership degree of element data to the working condition is determined by the cloud model,then it is converted into evidence.Based on evidence theory improved with similarity,the fusion of evidence is realized,and the evaluation of flotation working condition at the time node is completed.A qualitative trend analysis method based on an adaptive sliding window is used for the evaluation values sequence of working conditions to identify the flotation operation trend within the time interval.The validity of the method for evaluating the operation situation is verified by using industrial data in the flotation process.(2)To predict the future operation situation of the flotation process,a method based on time series information granulation is proposed to solve the poor prediction accuracy,caused by the accumulative error in the process of forecasting the operation situation.The Gaussian linear fuzzy information granules that describe the temporal and spatial trends well are used to granulate the original working condition evaluation value sequence.Taking the sequence of information particles as input,the NARX network is used to predict the future operation situation of the flotation process reflected by the information particle.The experimental results show that the prediction method can accurately predict the future operation situation of the flotation process.(3)A predictive control method of the flotation reagent based on the operation situation in the flotation process is proposed to solve the problem that it is difficult to make the flotation process run in the best situation under the action of the flotation reagent.At the level of information granule,combined with the steps of the predictive control algorithm,a model predictive controller is designed to make the flotation process run stably by adjusting the flotation reagent dosage.The stability of the control model is proved and the effectiveness of the control method is verified on the simulation platform.