Abstract: Flying-V layout is a classic non-traditional warehouse layout. In view of the storage location assignment problem with Flying-V layout, inventory efficiency and barycenter of storage goods are taken as the optimization objective based on its characters. A multi-objective optimization model of the storage location assignment is established. Subsequently, an adaptive genetic algorithm and a particle swarm optimizationare designed to solve the above issues. To accelerate the convergence and solve the premature problem of GA and PSO, adaptive strategies are adopted in the selection, crossover and mutation of the GA algorithm, and the inertia weight linear decreasing strategy is designed in PSO, which enhances the optimization performance of the algorithm. The genetic operators aredesigned, and the specific encoding is given. Finally, to verify the effectiveness and superiority of the proposed adaptive GA and PSO algorithm, a case is implemented with MATLAB software. Compared with adaptive GA algorithm, the results demonstrate that the proposed PSO algorithm has superior performance both in convergence rate and optimization effect. Contributions of the paper are the modeling and solving of the storage location assignment based on the Flying-V warehouse layout.

Key words: non-traditional layout warehouse, flying-V layout, storage-location assignment, adaptive genetic algorithm, particle swarm optimization

摘要： Flying-V是一种典型的非传统布局方式,根据其布局方式的特性,针对仓储货位分配优化问题,以货物出入库效率最高和货物存放的重心最低为优化目标,建立了货位分配多目标优化模型,并采用自适应策略的遗传算法(GA),以及粒子群算法(PSO)进行求解。根据货位分配的优化特点,在GA算法的选择、交叉和变异环节均采用自适应策略, 同时采用惯性权重线性递减的方法设计了PSO算法,有效地解决了两种算法收敛速度慢和易“早熟”的问题,提高了算法的寻优性能。为了更好地表现两种优化求解算法的有效性和优越性,结合具体的货位分配实例利用MATLAB软件编程实现。通过对比分析优化结果表明,PSO算法在收敛速度和优化效果方面相比于自适应GA算法更具有优势,更加合适于解决Flying-V型仓储布局货位分配优化问题。

关键词: 非传统仓储, Flying-V布局, 货位分配, 自适应遗传算法, 粒子群算法