Computer aided process planning(CAPP) is an important content of computer integrated manufacturing,and intelligentizing is the orientation of development of CAPP. Process planning has characters of empirical and time-consuming to finalize,and the same technical aim always can be achieved by different process schemes,so intelligentizing of process decision making always be a difficult point of CAPP and computer integrated manufacturing (CIM). For the purpose of intelligent aided process decision making and reuse of process resource,this paper proposed a decision making method based on rough sets(RS) and regular distance computing. The main contents and methods of process planning decision making are analyzed under agile response manufacturing environment,the concept of process knowledge granule is represented,and the methods of process knowledge granule partitioning and granularity analysis are put forward. Based on the theory of RS and combined the method of process attributes importance identification,the paper brought forward a computing model for process scheme regulation distance under the same attribute conditions,and conflict resolution strategy was introduced to acquire process scheme fit for actual situation of enterprise's manufacturing resources,so as to realize process resources' conflict resolution and quick excavate and reuse of enterprises' existing process knowledge,to advance measures of process decision making and improve the rationality and capability of agile response of process planning.
Configuration knowledge is a dynamic information set which is evolving and enriching on and on. Product model is the instantiation of configuration knowledge and the evolution of configuration knowledge is the essential inherent reason which causes the models dynamic evolvement. In the traditional model evolvement process, the inheriting and reuse of configuration knowledge was always ignored. Aim at solving the above problem, the multistage rhombus evolution mode of configuration knowledge is discussed in this paper. The product model based on configuration knowledge is put forward in different levels to achieve the models dynamic evolvement and automatic upgrading. The evolving configuration knowledge drives the product model to evolve directly according to the rule of up-layer evolvement. Furthermore, a new configuration knowledge reuse and optimization technology is presented to inheriting and reuse the foregone configuration knowledge in the course of model evolvement. At last, the air separation equipment which is related with the project is taken as an example to illuminate that the presented model evolvement and configuration knowledge reuse technology are validity and practical.
为准确地进行定量预测,提出了一种将仿真分析和集成径向基网络模型结合起来的制造系统性能指标预测方法。在定义和量化制造系统各类性能指标的基础上,分析了影响这些指标的静态和动态因素,并建立起径向基集成网络预测模型。通过基于Si mul 8平台的仿真分析来收集样本数据,最终利用Bagging方法训练出集成神经网络,实现对工件平均完工时间和设备利用率等关键性能指标值的预测。试验结果表明,采用该方法输入动态影响因素的取值后,能快速获得比较理想的性能指标预测结果,并且其预测精度明显高于其他的神经网络方法。
To illuminate the necessity of model evolvement and rouse,dynamics of injection molding machine' sproduct models are analyzed.The performance knowledge is used to support the model evolvement andreuse.The driven factors of mechanical product model are concluded.The dynamic characteristics ofproduct model are described.The.performance knowledge is used to improve specific evolvement process.The upper-layer passing rules are adopted in the mechanical product comfiguration design.The rules ofproduct model evolvement are investigated.And the model evolvement of injection molding machine hasthree levels.Practical and effective realization arithmetic is given to roalize the performance knowledgereuse.Finally,HT1800X1N series injection molding machines are taken as examples to illuminate thatthe arithmetic is correct and practical.
