Research On Sintering Burden Optimization Of Chongqing Iron&Steel Co Ltd.

The thesis do the researches on sintering optimization with the modern algorithms like Linear Programming, Improved Linear Programming, Monte Carlo, Genetic Algorithms, and then design four burden projects. After doing the experiment with the sintering tube, the chemical compositions, rotor index, and reduction ratio were measured. The feasibility of projects' got after analyzing the experiment data and the present states of CQISCO. All the work prepares for the directing the purchasing of raw and production of CQISCO to realize the low cost and good quality.The researches indicate:1 The projects through Monte Carlo and Genetic Algorithm agree with the command on chemical compositions, physical property, and metallurgy property, approve the quality, and reduce the cost. Comparing with the current project, the new projects change little, having the strong feasibility. The Linear Programming only consider the cost, the quantity of Australian ore is too high to order. So the feasibility is worse. Improved Linear Programming get the better level on rotor index and chemical compositions, but the velocity of reduction is low. A further research is needed.2 The relation from the point of view of burden among the chemical compositions, physical property, and metallurgy property is studied with step by step-regression. The relation is put into the burden optimization as the restriction condition. The fact illustrates the method effective.3 Solving the sintering burden optimization problem, once the restrict number is more, and the description of restrict is complex, Monte Carlo and Genetic Algorithm will be more effective. Due to the Penalizing Function of Genetic Algorithm, the restrictions can order by the importance. The Genetic Algorithm also has the better convergence velocity.