动态作业车间调度知识推理及知识系统设计

发布时间：2018-06-29 21:53

本文选题：动态环境 + 调度规则　；参考：《合肥工业大学》2017年硕士论文

【摘要】：在现代制造模式下,静态的调度方案已经无法适应于多变的作业车间生产环境,基于知识推理的调度方法是解决该类问题的有效方式之一。当前调度知识系统多建立领域内专家经验基础之上,多存在主观性强、决策依赖部分属性、多源知识冲突和知识滞后等问题。本文针对动态车间环境下的调度知识推理研究现状,重点研究调度规则的动态选择的问题,利用企业制造系统中的生产调度数据,运用遗传算法和BP人工神经网络算法构建动态调度知识网络,设计基于动态知识网络的调度知识系统。首先,建立静态作业车间数学模型;针对一般作业车间静态调度问题,通过编码解码、交叉、变异等遗传算法操作,获得问题最优解;归纳一般静态车间调度问题求解的遗传算法流程。其次,分析常见的作业调度规则和几种复合规则调度方式,确定本文研究方向为自适应调度;利用遗传算法求解改进的4×3调度问题的最优解,基于BP人工神经网络算法建模,定义网络输入参数和输出参数,从遗传算法最优解中抽取冲突时间决策点,计算人工神经网络输入和输出,获得训练样本;训练样本数据获得非线性网络关系,指导不确定车间条件下调度规则的选择。最后,分析调度知识系统实现的关键策略,进行调度知识系统的系统需求分析,归纳总结系统的业务流程;提出调度知识系统的硬件框架和软件框架,实现系统的关键数据库设计和软件模块设计。
[Abstract]:In the modern manufacturing mode, the static scheduling scheme can no longer adapt to the changeable job shop production environment. The scheduling method based on knowledge reasoning is one of the effective ways to solve this kind of problem. At present, most scheduling knowledge systems are based on the experience of experts in the field, and there are many problems, such as strong subjectivity, partial attribute of decision dependence, multi-source knowledge conflict and knowledge lag, etc. Aiming at the present situation of scheduling knowledge reasoning in dynamic workshop environment, this paper focuses on the dynamic selection of scheduling rules, and makes use of the production scheduling data in enterprise manufacturing systems. Genetic algorithm and BP artificial neural network algorithm are used to construct dynamic scheduling knowledge network, and a scheduling knowledge system based on dynamic knowledge network is designed. Firstly, the mathematical model of static job shop is established, and the optimal solution of the problem is obtained by genetic algorithm, such as coding and decoding, crossover, mutation and so on. The genetic algorithm flow of general static job shop scheduling problem is summarized. Secondly, by analyzing the common job scheduling rules and several complex rule scheduling methods, the research direction of this paper is determined as adaptive scheduling, the genetic algorithm is used to solve the optimal solution of the improved 4 脳 3 scheduling problem, and the BP artificial neural network algorithm is used to model the model. The input and output parameters of the network are defined, the conflict time decision points are extracted from the optimal solution of genetic algorithm, the input and output of artificial neural network are calculated, the training sample is obtained, and the nonlinear network relation is obtained from the training sample data. To guide the selection of scheduling rules under uncertain job shop conditions. Finally, the key strategies of scheduling knowledge system are analyzed, the system requirements of scheduling knowledge system are analyzed, the business process of the system is summarized, and the hardware and software framework of scheduling knowledge system is put forward. The key database design and software module design of the system are realized.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TB497

【参考文献】