将热轧批量计划编制问题归结为奖金收集的车辆路径问题,按照热轧工艺规程,同时考虑轧制计划中钢板宽度、厚度的反跳约束,设计了反跳惩罚表,提出一种多目标数学规划模型,即最小化轧制计划数、最小化惩罚值、最小化未轧制板坯数.应用粒子群(PSO)算法进行求解,对PSO算法的求解过程进行了改进,使其适用于热轧批量计划问题,在实验中测试了算法的参数(惯性权重、加速因子)对解的影响,并进行了实验分析,获得了满意的结果.实验表明,经过改进的PSO算法所获得的最优解以及平均使用时间上都是有效的.
This paper formulates the hot-milling batch planning problem as a prize collecting vehicle routing problem(PCVRP).According to the hot-milling technics rules,this paper considers the inverse bounce of the width and the thickness of the steel strips and presents a multi-objective mathematics programming model.The objectives include minimizing the number of hot-milling plans,minimizing the total penalties,and minimizing the number of slabs without being milled.To solve this problem,an improved particle swarm ...