不确定条件下装配式建筑生产‒运输集成调度双层优化

Bi-Level Optimization of Integrated Production-Transportation Scheduling in Prefabricated Construction Under Uncertainty

摘要:

从构件供应商角度出发，研究了不确定条件下装配式 建筑生产?运输集成调度，建立了以成本?满意度为目标的多流水线双层优化模型。采用模糊数描述构件生产加工时间和运输时间、安装时间窗的不确定性。在上层模型中，以多条流水线生产?运输总成本最小为目标，采用遗传算法进行订单流水线分配；在下层模型中，以模糊悲观准则确定满意度最大为目标，采用自适应贪婪禁忌遗传算法（AGTGA）进行单条流水线生产?运输集成调度。上层决策方案与下层决策方案不断迭代实现不确定条件下装配式建筑生产?运输集成调度优化方案。结果表明，AGTGA相比于遗传算法、迭代贪婪遗传算法表现出较好的性能，能实现在不确定条件下上下层模型的最优目标，且能确定装配式建筑生产?运输集成调度的最优方案。

Abstract:

This paper， from the perspective of component suppliers， investigates integrated production-transport scheduling of prefabricated construction under uncertainty. A bi-level optimization model， considering cost and satisfaction objectives， is developed to describe the uncertainty of component production processing time， transport time， and installation time windows with fuzzy numbers. In the upper-level model， a genetic algorithm （GA） is employed to determine the optimal order allocation among production lines to minimize the total production-transportation costs. In the lower-level model， the adaptive greedy tabu genetic algorithm （AGTGA） is proposed to optimize the integrated production-transportation scheduling on a single pipeline production-transportation integrated scheduling to maximize satisfaction based on fuzzy pessimistic criteria. Besides， an iterative process achieves optimal scheduling solutions under uncertainty by exchanging solutions between the upper and lower levels. The results indicate that AGTGA outperforms iterated greedy algorithm （IGA） and GA， achieving optimal objectives of upper and lower models under uncertainty and determining the optimal solution for production-transport integrated scheduling of prefabricated buildings.