Research Initiation Award: Automatic Precedence Knowledge Acquisition and Design of a Real-Time Intelligent Assembly Scheduler with Neural Computing

本研究的目的是开发一个从装配设计过程到最优装配计划生成的实时智能装配系统。研究的目标是：(1)从基于规则的系统或神经网络学习方案中自动获取装配对象的优先知识，(2)找到优先知识与被调查神经网络之间的映射，(3)基于优先知识与网络和算法之间的映射，生成最优装配调度；(4)实现了集成设计过程和最优装配计划生成的系统智能装配系统。该系统适用于汽车装配、电子线路板装配、工业柔性制造等相关领域的优先约束计划调度问题。

The objective of this research is to develop a real-time intelligent assembly system which integrates the operations from the assembly design process to the optimal assembly schedule generation. The goals of the research are: (1) obtain the precedence knowledge of the assembled object automatically either from a rule-based system, or from neural network learning schemes; (2) find the mapping between the precedence knowledge and investigated neural networks; (3) generate the optimal assembly schedule based on the mapping between the precedence knowledge and the network and algorithms; and (4) implement a systematic intelligent assembly system which integrates the design process and the generation of the optimal assembly schedule. The system is applicable to automobile assembly, electronic circuit board assembly, industrial flexible manufacturing and other fields related to the precedence-constraint planning and scheduling problems.