响应动态约束条件的多目标货位优化算法研究

针对自动化立库货位决策与优化问题，考虑到优化目标多样、托盘使用状态及可分配货位动态变化等因素，提出了一种响应动态约束条件的多目标货位优化算法。以巷道作业均衡、货架重心稳定及作业路径最短建立多目标优化模型，基于变异系数自适应差分进化算法，使用货位随机数编码，根据实时货位可行域进行个体解码，以响应动态货位约束条件。提出了基于层次分析的Pareto解评价方法，从而获得多批作业货位持续优化的目标权重，为仓储货位决策提供合理方案。多批作业算法实验结果表明:所提算法效果显著优于多目标简单加权算法，能够有效应用于动态货位决策与优化。

Multi-objective location optimization algorithm in response to dynamic constraints

Considering the storage location decision and optimization problems in automated storage and retrieval system, we propose a multi-objective logistics optimization algorithm, which considers various optimization objectives, such as the usage status of the pallet and dynamic changes in the allocable storage location. A multi-objective optimization model is established based on the equilibrium of roadway operations, the stability of the gravity center of shelves, and the shortest operation path. Based on the adaptive variation coefficients’ differential evolution algorithm, a random number encoding of the storage location is used to perform individual decoding according to the real-time feasible domain in response to the dynamic constraint condition. A Pareto optimal solution evaluation method based on the analytic hierarchy process is proposed to obtain the target weight related to the continuous optimization of a multi-batch operation, and a reasonable plan for the storage location decision is provided. The experimental results of the multi-batch operation show that the proposed algorithm is significantly better than the simple weighting algorithm, which can be effectively applied to dynamic location decision and optimization.