基于案例推理的工业生产过程优化方法的研究

发布时间：2018-03-23 06:26

本文选题：过程优化　切入点：案例推理　出处：《东北大学》2014年硕士论文　论文类型：学位论文

【摘要】：过程优化可使企业在原有设备条件的基础上,提高生产效率,满足节能减排的要求,获得显著的经济效益,研究工业生产过程优化的方法既有理论意义,又有实用价值。本文提出了一种基于案例推理技术的工业生产过程优化方法。完成了对案例库的构造、扩充和推理决策准确性、快速性的改进,实现了案例推理在线解决过程优化问题的方法。论文主要工作如下：首先,分析了案例推理的原理,针对案例推理技术进行过程优化时存在的问题进行改进,运用遗传算法对案例库进行约减,同时优化设定案例特征属性的权值,提高案例推理的速度和决策精度；运用蚁群聚类算法建立案例库两级索引,减少案例检索时遍历案例的数量,提高检索速度。改进后的案例推理技术能更精确迅速地进行过程优化决策。然后,提出一种应用过程优化模型构造源案例的方法。在该方法中,采用本文提出的输入变量取值策略,对每一组输入取僮由基于分离比较策略的粒子群算法对优化模型进行求解,得到结合经验构造的过程优化源案例,进而通过案例学习策略构建案例库。以此方法构造源案例,丰富了推理决策的案例库,保证了案例推理的求解质量和决策精度。最后,提出一种基于案例推理的工业生产过程优化方法的框架,该框架包含了采用案例推理技术进行一般工业生产过程优化决策的基本流程和各个步骤。在前两步离线准备完成的基础上,设计了在线应用阶段各个环节的实现方法。将该方法应用到球团配料生产过程中,取得了良好的效果。
[Abstract]:Process optimization can make enterprises improve production efficiency, meet the requirements of energy saving and emission reduction, and obtain remarkable economic benefits on the basis of the original equipment conditions. It is of theoretical significance to study the method of industrial production process optimization. This paper presents an optimization method of industrial production process based on Case-Based reasoning (CBR), which improves the construction of case base, expansion of case base, accuracy and rapidity of reasoning decision. The main work of this paper is as follows: firstly, the principle of CBR is analyzed, and the existing problems in Case-Based reasoning (CBR) are improved. Genetic algorithm is used to reduce the case base, at the same time, the weight value of the case characteristic attribute is optimized to improve the speed of case-based reasoning and the decision accuracy, and the ant colony clustering algorithm is used to establish the two-level index of the case base. Reduce the number of cases traversed in case retrieval, improve the retrieval speed. The improved case-based reasoning technology can make process optimization decisions more accurately and quickly. Then, A method of constructing source cases by using process optimization model is proposed. In this method, each group of input fetches is solved by particle swarm optimization algorithm based on separation and comparison strategy, using the input variable selection strategy proposed in this paper. The case base is constructed by case learning strategy, which enriches the case base of reasoning decision, and ensures the solution quality and decision accuracy of case-based reasoning. A framework of Case-Based reasoning (CBR) based optimization method for industrial production process is proposed. The framework includes the basic process and each step of the general industrial production process optimization decision based on the case-based reasoning technology. On the basis of the first two off-line preparation, The realization method of each link in the online application stage is designed, and the method is applied to the pellet batching production process, and good results are obtained.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TB49

【相似文献】