The purpose of this paper is to optimize dynamic scheduling in steel-making, the main case adaptation approaches and the case adaptation framework in case-based reasoning were summarized. Based on that, the Poliscastro reconstruction method was applied, the adaptation knowledge was obtained by using the target case and similar cases in the input and output, an improved portfolio adjustment algorithm was proposed for the complex case adaptation problems. Finally, using the improved adjustment technology, the dynamic scheduling of steel production is solved. The results show that the proposed technique is effective and accurate in solving the complex and changeable case adjustment problem.