Research On Hybrid Modeling Method Of Copper Extraction Process In Hydrometallurgy

The high-quality mineral resources that have been explored are nearing poverty,and the main raw materials are gradually shifting to complex low-grade ore resources or secondary resources that are "difficult to survey,difficult to mine,and difficult to process." Fall into the danger of survival competition.The change of main raw materials and the implementation of sustainable development strategy and the requirement of advocating green industry to protect the ecological environment,the hydrometallurgy technology as one of the two major extraction technologies is more advantageous.The mature,low-cost,low-risk solvent extraction makes the hydrometallurgy technology develop more rapidly.The new extractant and extraction equipment improve the performance of solvent extraction,but the component concentration measurement still stays in the offline laboratory analysis stage,which cannot reflect the level of metal separation in time.The currently established mathematical model is not completely suitable for control System analysis and design.Therefore,the lack of dynamic models suitable for industrial applications has hindered the development of advanced process control systems for copper solvent extraction processes.This thesis aims at the problem that the concentration of components in the hydrometallurgical copper extraction production process is difficult to measure online.Based on an in-depth understanding of the process characteristics of the copper extraction process,a hybrid modeling method combining a mechanism model and a data model is used to analyze the wetness suitable for industrial applications.The dynamic model of the copper extraction process in the process metallurgy has been comprehensively and systematically studied.The main researches are summarized as follows:(1)On the basis of in-depth understanding of the hydrometallurgical copper extraction process technology principles,the dynamic mechanism model of the copper extraction process is established according to the conservation of materials.The model is composed of the aqueous phase copper conservation equation,the organic phase copper conservation equation and the equilibrium of extraction equilibrium The composition of the nonlinear relationship of concentration,where the equilibrium concentration is estimated using McCabe-Thiele plots,the copper conservation equation of the aqueous phase and the organic phase contains the copper ion kinetic reaction mass transfer rate model,and the factors affecting the extraction equilibrium process are simulated by simulation analysis.(2)The unknown parameter kinetic reaction mass transfer rate in the copper extraction process mechanism model is not measurable in practice.In this thesis,the finite difference,polynomial fitting,Tikhonov regularization and wavelet analysis methods for copper extraction process kinetic reaction mass transfer rate estimation strategy,Respectively,to estimate the mass transfer rate of the kinetic reaction in two cases with noise and without noise.The simulation results show that the Tikhonov regularization method can obtain satisfactory and acceptable estimation results in both no-noise and noisy situations.This method minimizes the impact of measurement noise in the concentration measurement data on the mass transfer rate estimation.This will lay an important foundation for the establishment of subsequent data models,production index prediction and operation optimization.(3)Establish a serial mixing model of the copper extraction process,which consists of a mechanism model with unknown parameters(mass transfer rate)and a data estimation model with unknown parameters in series.Based on the results of Tikhonov regularization estimation as training data,the BP neural network and the kernel partial least squares(KPLS)were used to establish data estimation models for the kinetic reaction mass transfer rate,and then the established unknown parameter estimation models were connected in series with the mechanism model.Into a mixed model,and finally the effectiveness of the mixed model is confirmed by simulation experiments on the copper ion concentration of the raffinate.