Predictive Control Of Reagent-addition Amount Based On Distribution Of Bubble Size In Copper Roughing Flotation Process

The flotation process is a complex physical and chemical process completed in the gas, solid and liquid phases, so it is difficult to establish an effective mechanism model. And it is difficult to achieve the automatic control of reagent dosage in the lack of direct evaluation criterion as the reference. Currently, the reagent dosage is still controlled mainly by manually observing the froth status, which leads to large reagent consumption and unstable flotation operation conditions. The bubble size distribution is closely related to the performance of the flotation in the mineral flotation process, and flotation reagents addition amount will directly affect the bubble size distribution. Therefore, the optimal dosage control based on bubble size distribution characteristics is of very realistic significance for stabilizing the flotation working condition and increasing the flotation efficiency.In view of the bubble size distribution characteristics and the fact that it is difficult to represent the surface bubble size distribution features of different types of froth images in the numerical representation using the conventional methods, a method based on the minimum entropy principle where B-spline is used as the basis function for probability density estimation (PDF) is proposed in this paper. In this method, the third-order B-spline basis function polynomial is adopted to fit the probability density function of the bubble size, and the minimum information entropy principles is utilized to set the optimal number estimation of B-spline basis functions, finally, the linear combination weights are used to characterize the bubble size distribution features.In order to optimize the reagent-addition operation in the flotation process, make the flotation state stable and achieve the optimal reagent-addition amount control, a reagent dosage predictive control method based on the probability density function of bubble size is put forward in this paper. In this method, the PDF model of bubble size is built. With the aim to control flotation bubble size distribution, the generalized predictive control algorithm is utilized to approach target bubble size PDF in order to achieve the optimal reagent dosage control. For the constraint problems in the system, the method based on the combination of genetic algorithm and nonlinear programming (GA-NP) is adopted to solve the optimal control effect instead of the rolling optimization strategy used in the conventional prediction model.The proposed algorithm is verified through the simulation experiments on industrial field data in this paper, the experimental results demonstrate that this method can acquire the bubble size distribution characteristics simply and conveniently, and it can effectively track the bubble size distribution features during the reagent addition operation in the copper roughing flotation process by using the offline data identification model. There are43figures,3tables and67references in this dissertation.