基于三阶段混合启发式算法的带纹理板材“一刀切”下料问题研究

Research on the “One-Cut” Panel Cutting Problem with Textured Panels Based on a Three-Stage Hybrid Heuristic Algorithm

  • 摘要: 本研究将三阶段的非精确定位算法与遗传退火(GASA)算法结合来解决家具制造行业的标准尺寸板材考虑纹理工艺约束的矩形毛坯板下料问题。由于需要根据家具制造生产的实际情况来实现基于锯切要求下的“一刀切”生产且考虑板材与毛坯纹理问题的排样问题优化,本文针对现有家具生产的大规模套裁排样方案毛坯板材多、尺寸丰富、板材纹理导致毛坯方向不可旋转等约束,给出详细的三阶段非精确的定位算法结合在两点变异后加入退火降温操作的遗传退火的元启发式算法,形成一种改进的混合启发式算法;通过经典算例测试和实际生产示例的实验验证,结果表明,本研究提出的三阶段混合启发式算法在不同批次下料问题中其板材利用率均优于单一定位策略算法,且三阶段非精确混合启发式算法在考虑纹理方向的多种类套裁排样问题中相较于三阶段非精确算法、两阶段启发式算法及三阶段匀质混合启发式算法利用率多算例下平均提高5.98%、5.09%、3.31%,从无纹理下料生产场景切换考虑纹理方向的下料生产场景时,三阶段匀质混合启发式算法的利用率平均降低速度约为该算法的3倍,证明了三阶段非精确混合启发式算法在“一刀切”各类板材下料生产场景下具有广泛适用性与稳定性。

     

    Abstract: This study combines a three-stage non-exact positioning algorithm with a Genetic Annealing (GASA) algorithm to address the rectangular blank layout problem for standard-sized textured panels in the furniture manufacturing industry under process constraints. To optimize the "one-cut" production strategy considering sawing requirements and panel texture alignment, we propose an improved hybrid heuristic algorithm. This algorithm integrates a three-stage non-exact positioning method with a metaheuristic GASA algorithm enhanced by two-point mutation and simulated annealing cooling. It specifically tackles constraints such as large-scale nested layouts, diverse blank sizes, and non-rotational texture alignment in furniture production. Experimental validation using benchmark cases and real-world production scenarios demonstrates that the proposed three-stage hybrid heuristic algorithm achieves higher material utilization rates compared to single-stage positioning strategies. In multi-category nested layout problems with texture constraints, the algorithm outperforms three-stage non-exact algorithms, two-stage heuristic algorithms, and homogeneous three-stage hybrid algorithms by average utilization improvements of 5.98%, 5.09%, and 3.31%, respectively. Furthermore, when transitioning from non-textured to textured cutting scenarios, the homogeneous three-stage hybrid algorithm exhibits a material utilization decline rate approximately three times faster than the proposed method. These results confirm the broad applicability and stability of the three-stage non-exact hybrid heuristic algorithm across diverse "one-cut" panel cutting scenarios.

     

/

返回文章
返回