改进粒子群算法的动态空间调度方法

针对船体分段生产调度的多目标性和动态性,提出了一种改进粒子群算法的动态空间调度方法,确定船体分段在工作平台上的加工顺序和空间布局位置.算法以加工完成时间最短和空间利用率最高为目标,采用自适应惯性权重策略保证算法的收敛性,并引入遗传算法中的选择算子和变异算子增强算法的收敛速度和多样性,利用启发式定位策略确定分段的位置.最后,以船厂实际生产数据进行仿真验证.仿真结果表明,所提方法可以大大降低以手工方式制定调度计划的复杂度,并能有效地提高空间利用率达到70%,说明该方法是解决动态空间调度问题的一种有效方案.

Dynamic spatial scheduling approach based on improved particle swarm optimization algorithm

Multiple objectives and changing needs for the use of space in shop-floor scheduling systems used by shipbuilding yards make planning when and where to start block construction became a challenge. A block includes multiple steel plates in sections of various sizes and shapes and competes for time and space with all other potential blocks. A dynamic spatial scheduling approach based on an improved particle swarm optimization algorithm was proposed to determine the optimal processing sequence and spatial location of blocks. To minimize processing time and maximize spatial utilization, the adaptive inertia weight strategy was used to ensure the algorithm converged. A selection operator and mutation operator were used in the algorithm to improve the convergence rate and prevent a local optimum. A heuristic location strategy was developed to determine the location of blocks. The simulation results indicate that the proposed algorithm can simplify the scheduling making, and improve the spatial utilization to 70%, explaining that the proposed algorithm is an effective plan to solve the dynamic spatial scheduling.