半导体封装阶段批加工设备调度研究
文献类型：学位论文
| 作者 | 张奎
|
| 学位类别 | 硕士 |
| 答辩日期 | 2016-05-25 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 导师 | 刘昶 |
| 关键词 | 半导体封装测试 分布估计算法 概率模型 启发式规则 批调度 |
| 其他题名 | Research on Batch Scheduling in Semiconductor Assembly Manufacturing |
| 学位专业 | 控制工程 |
| 中文摘要 | 本文以半导体封装阶段批加工设备调度研究为背景，分别针对以最小化最大完工时间和最小化最大加权误工时间为性能指标的批加工设备调度，提出了一种基于r-P规则的EDA算法和一种组合优化算法（EDA-BATC），并通过仿真测试，验证了各个新算法的有效性和可行性。主要研究内容和取得的成果如下：1）以半导体生产制造过程中的封装测试阶段为背景，分析了其生产制造过程中独有的特点及工艺制造流程，针对装片和塑封中固化环节的批加工设备调度给出问题描述，并分别针对以最小化最大完工时间和最小化最大加权误工时间为性能指标的批加工设备调度分别展开了相关算法研究。2）针对半导体封装测试阶段中以最大完工时间为性能指标的批加工设备调度问题，提出了一种新型基于r-P规则的EDA算法，该算法的主要思想是：通过减小组批等待时间和加工时间较长的工件优先入批来使机器空闲时间最小化，并结合EDA算法的进化思想，最后实现优化最大完工时间的目的。3）设计了一种组合优化算法（EDA-BATC）来处理以最小化总加权误工时间为性能指标的批加工设备调度问题。首先通过实验设计证明了EDA算法在处理最小化总加权误工时间的单机批调度问题上的有效性。随后将多机批调度问题拆分为分批与批调度两个子问题，通过组合算法的设计，充分考虑了时间复杂度与求解质量，通过与其它算法的对比，充分体现了该算法的优越性。 |
| 英文摘要 | In this paper, aiming at minimizing the makespan and the maximum weighted idle time for batch scheduling of the processing equipment in the semiconductor packaging stage, we proposed a rule-based rP EDA algorithms and one combinatorial optimization algorithms (EDA-BATC), and we verify the effectiveness and feasibility of each new algorithm through the simulation tests . The main contents and results achieved are as follows: 1) In the packaging and testing stage of the semiconductor manufacturing process, we have analyzed its manufacturing process and the unique characteristics of the manufacturing process, and give the problem description against the batch scheduling of the processing equipment in the cured part of loading and plastic, we also carry out studies of related algorithm about aiming at minimizing the make span and the maximum weighted idle time for batch scheduling of the processing equipment. 2)In the semiconductor assembling and testing phase ,aiming at minimizing the make span and the maximum weighted idle time for batch scheduling of the processing equipment, we have proposed a new EDA algorithm based on rP rule, the main idea of this algorithm is minimizing the machine idle time by reducing the waiting time and processing the job with the longer processing time with the priority, meantime, combining the evolutionary thought in the EDA algorithm, and finally achieving the optimization in the make span. 3) we have designed a combinatorial optimization algorithm (EDA-BATC) to deal with the problem of minimizing the total weight loss of working time in the batch processing equipment scheduling. Firstly, through experiments designing, we proved the effection of the EDA algorithm in dealing with total weight loss of working time in the single batch scheduling problem. Then we split the batch scheduling problem into two sub-problems: batch operation and batch scheduling In the combinatorial optimization algorithm (EDA-BATC), we have considered the time complexity and quality of solutions. At last, through the comparison with other algorithms, we embodies the advantages of this algorithm. |
| 语种 | 中文 |
| 产权排序 | 1 |
| 页码 | 60页 |
| 源URL | [http://ir.sia.cn/handle/173321/19630]  |
| 专题 | 沈阳自动化研究所_数字工厂研究室 |
| 推荐引用方式 GB/T 7714 | 张奎. 半导体封装阶段批加工设备调度研究[D]. 中国科学院沈阳自动化研究所. 2016. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。