改进NSGA2在炼钢-连铸调度中的应用研究

Applied Research for Steelmaking and Continuous Casting Production Scheduling Based on Improved NSGA2

针对具有多重精炼方式的炼钢-连铸调度排产问题,采用单纯带精英策略的非支配排序遗传算法(NSGA2)存在求解精度不高,求解速度慢、并得到Pareto解集后需要人工确定最优解的问题,提出了基于优先级策略的改进NSGA2新方法。首先基于炼钢-连铸生产工艺过程及调度目标和要素建立多目标优化调度模型,然后将所提出的方法应用到具有多重精炼的炼钢-连铸生产调度问题中,并与现有采用原始NSGA2的仿真结果进行比较。实验结果表明本文提出的基于优先级策略的改进NSGA2算法在求解速度、求解精度上均优于原始NSGA2算法,并能自动给出唯一的最优调度方案,避免人工确定最优解缺乏科学依据问题。

In this paper a new improved non-dominated sorting genetic algorithm with elite strategy(NSGA2)based production scheduling method is proposed to solve problems that the solving precision is not high,the solving speed is slow and the optimal solution needs to be determined artificially after the Pareto solution set is obtained,when using the traditional NSGA2 to solve the scheduling problem of steelmaking-continuous casting complicated production process with multiple refining methods.At first,the multi-objective optimization scheduling model is established according to production process and schedule requirement,and then the method proposed in this paper is applied to solve the production scheduling problem of steelmaking and continuous casting with multiple refining.Experimental results show that compared with the original algorithm,the improved algorithm can solve the problem faster and find more accurate solutions.By the using the improved algorithm,the only optimal scheduling scheme can be automatically given,which avoids the lack of scientific basis for the artificial determination of the optimal solution.