Lead-zinc Sintering Process Optimization For The Integrated Production Goals

Lead-Zinc sintering process is a step of preparing the raw materials in the reduction sintering of Imperial Smelting Process. In order to meet the raw materials requirements of the reduction smelting blast furnace, the ultimate goal of sintering production is to produce the sintering agglomerate which has the right chemical composition, a certain mechanical strength and reducibility. The production and quality of the sintering agglomerate have decisive influence on the productivity of Lead-zinc smelting process.By adjusting the operating parameters, comprehensive production index of the sintering process is bound for ensuring the quality of sintered block, raising the output of sintering agglomerate, and making the sintering state stable, so that with the optimal operations sintering process could achieve high yield, good quality, stability requirements. Burning through point (BTP) is an important parameter of the state of sintering process, which is a main indicator of good or bad. In order to meet the requirements of comprehensive production index, the optimization of sintering process will be divided into the production optimization of sintering agglomerate and BTP optimization. Based on the deep analysis of the lead-zinc sintering process, the optimization basic idea of the sintering process on comprehensive production index is presented and the optimization methods of sintering agglomerate production and quality are proposed.First of all, according to the mechanism analysis and artificial experience, the impact factors of the comprehensive production index are discussed by qualitative analysis, and by using gray relational analysis method the main impact factors of the comprehensive production index have been attained through quantitative calculation. Then, based on the mechanism analysis and gray relational analysis, the prediction models of sintering agglomerate production and quality have been established by making use of improved BP neural network. In accordance with the detection method of BTP, integrated prediction model have been established based on the historical information and the gray system theory.Secondly, according to the prediction models of sintering agglomerate production and quality, the optimization model have been established, in which the decision-making variable operating parameters are constrained by the quality requirements and process conditions, for the optimization model of production quality, Then, by using the improved SPEA2-based particle swarm optimization algorithm, the optimization set values of decision-making variables are obtained. On the other hand, according to the BTP integrated prediction model, BTP optimization model for the objective optimizing settings has been established, in which the decision-making variable operating parameters are confined by the produce boundary conditions. The optimization set values of operation parameters are attained by the use of adaptive differential evolution algorithm based on the quadratic variation.Finally, in order to estimate the actual sintering process conditions, a comprehensive evaluation model of sintering conditions is established. On the base of the idea of weight entropy, the final optimization set values of operating parameters are achieved by the optimizing strategy of coordinate optimization. The simulation experiments show that the optimization methods oriented proposed in this dissertation the comprehensive production index have been obtained better optimal results, which set a vital foundation for the optimal operations of the sintering production.