Batch-to-Batch Optimization Of Cobalt Oxalate Synthesis Process Based On Modifier-Adaptation Scheme

With the development of manufacturing industry and high-technology industry,cobalt plays a more and more important role in spaceflight and aviation,hard alloy,and batteries production.Because of the wide application of cobalt powder in many high-tech industries,the demands of mean particle size of cobalt become higher,which rely heavily on the cobalt oxalate mean particle size.However,at present,the products quality control in the hydrometallurgical synthesis process of cobalt oxalate mainly depend on the worker’s experience,which make the mean particle sizes of cobalt oxalate are vary from batch to batch.In addition,the filter will be clogged easily and drying process will slow down when the mean particle size is too small,which will reduce the production efficiency of cobalt oxalate.So it is urgent to introduce advanced control technology to improve products quality.In the thesis,In order to improve the production efficiency of cobalt oxalate,a data based model is built,on the basis of systematically analysis of characters of cobalt oxalate production process and mechanism of cobalt oxalate synthesis.Finally,a batch-to-batch modifier-adaption optimization is used to find the optimal operation point of mean particle size of cobalt oxalate.The main researches are summarized as follows:(1)Introduce the production process of cobalt oxalate synthesis and qualitatively analyze the influence of different kind of manipulate variable on the mean particle size of cobalt oxalate.Additionally,a mechanism model is built according to applying crystallization kinetics,the population balance theory,and the mass balance theory.Finally,the simulation result imply that the qualitative analysis is corresponding with simulation.(2)The MPLS model is used to build the prediction model of mean particle size of cobalt oxalate,taking use of historical production data.By simulation experiment,the efficiency of data base modelling method is verified.(3)Due to the model uncertainty in cobalt oxalate synthesis process which will make the optimal control trajectory of model deviate from the real one of plant.In order to overcome the difficulty,a modifier-adaption optimization method is used which take advantage of an adaptive modification term to correct the objective function and the constraint conditions.This method make the iteration results converge to the practical optimal operating point.Finally,the modifier-adaptation optimization method is used in mean particle size optimization of cobalt oxalate to prove the efficiency of the method.(4)In order to decrease the error brought by model generalization and gradually expanding the region of iterative optimization,the modifier-adaption optimization is improved in two aspects.Firstly,constraints is used to confine the optimal solution in the valid region of the data-driven model and reduce the error of prediction.Secondly,the MPLS model is update during the iterative optimization to expand the searching region of optimization.By simulation experiment,the efficiency of the method was verified.