Research On Addition Optimization Model Of Basic Oxygen Furnace Steelmaking Based On Particle Swarm Optimization

As the last link of BOF (Basic Oxygen Furnace) steelmaking process, deoxidizing and alloying addition during tapping determines the quality of blowing. The decision of alloy addition is the primary means of control of the process, and its purpose is to enable the various elements in the ladle in order to achieve the production requirements while minimizing the cost of alloy, therefore, alloy addition optimization model is necessary. Traditional optimization algorithm is complicated and difficult to deal with complex objective function, so that the model can not meet the complexity of oxygen alloying the actual production. Aim at the problems above, this paper propose optimization model for alloy addition using dynamic neighborhood multi-objective particle swarm optimization (DN-MOPSO).Control process of BOF has high quality requirements of real-time optimization algorithm, this paper improves convergence speed and the quality of solution set of DN-MOPSO, to make it more suitable for alloy addition optimization model. First of all, in order to come over slow convergence of the original DN-MOPSO and the poor completeness of the solution set, this paper proposes extended external storage solution set, combined with the improved neighborhood selection strategy to improve algorithm optimization speed, while the optimal solution can obtain a complete set; In addition, for the uneven distribution of DNMOPSO solution set, a kind of solution set maintenance methods,namely adaptive grid algorithm, has been introduced into DN-MOPSO, the algorithm optimal set can be evenly distributed with a good diversity, to provide decision makers a more comprehensive solution; Finally, the improved multi-objective dynamic neighborhood particle swarm optimization is applied to the optimization model of BOF alloy addition, the actual production data has been used to simulate, and verify the excellent performance of the improved algorithm and the validity of the alloy addition optimization model. The model proposed also can direct the actual production of BOF.