| Summary: | 碩士 === 華梵大學 === 工業管理學系碩士班 === 93 === The research focuses on multiple-container packing problems with considerations of multiple constraints. The space utilization, stability, and loading sequence of objects are also considered in order to make results more practicable. The objective of the research multiple-container packing problem is to minimize container cost to pack all data objects under the conform to the considerations of stable packing, load bearing limitation, highly space utilization, and efficiently unhindered unloading. Therefore, clustering technology and genetic algorithm are combined to solve such complex multiple-container packing problems under these constraints.
At the beginning, clustering algorithm is applied to classify data objects into different groups with varied characteristics within. The grouping procedure is done by considering dimension of objects, unloading sequence of objects, and capacity of container. Then genetic algorithm combines with heuristic rules is used to pack data objects into containers. The stable packing, space utilization, unhindered unloading, and load bear limitation are the major considerations in this stage.
A computer system based on the proposed algorithm was developed. The computer system is not only a simulation tool for performance analysis, but also a system to provide practical solutions for customer designated multiple-container packing problems. Thousands of cases were simulated and analyzed to evaluate the performance of proposed research and prove the applicability in real world.
|
|---|
