Multi-source Data Fusion Research For Discrete Manufacturing Smart Workshops

Discrete manufacturing intelligent workshop is an important embodiment of intelligent manufacturing,it is based on the workshop data as the carrier,and through the advanced information technology,to realize the automatic decision-making and lean management.Data fusion is one of the most important technologies in the intelligent workshop,which can integrate and complement the workshop data,and provide a good technical support for the production and decision-making of the intelligent workshop.However,the existing data fusion methods are difficult to deal with the multi source heterogeneity,mass and uncertain data in discrete manufacturing intelligence workshop.In view of this,this paper puts forward the corresponding data fusion method for the discrete manufacturing intelligent workshop.The main research work includes:(1)This paper introduces the research background,current situation and the related concepts of the intelligent workshop and data fusion,and analyzes the data characteristics and the integration requirements of the discrete manufacturing intelligence workshop.(2)According to the data characteristics of multi source,heterogeneity and uncertainty in discrete manufacturing workshop,a new multi-source data fusion method is proposed.First of all,the cloud model parameters of each object are obtained,and then the BPA of each data corresponding to each object is calculated by cloud parameters.Finally,each BPA is fused into the corresponding knowledge under the condition of certain rules.The experimental results show that this method can effectively fuse the multi source heterogeneity and uncertain data.(3)According to the mass data characteristics of the discrete manufacturing intelligent workshop,and the problem that the existing fusion methods are difficult to deal with,a mass data fusion method based on clustering is proposed.This paper firstly uses the K-means clustering algorithm to processing workshop massive data into multiple clusters,and each cluster center as a representative information of the cluster;then weighting to represent information,and calculate the weighted information of BPA;Finally,using some methods to fuse BPA.The algorithm complexity and simulation results show that the proposed method can effectively fuse the mass data in the discrete manufacturing intelligent workshop.