Research on Data-driven Performance Degradation Mechanism Modelling and Predictive Control Method for Discrete Manufacturing System
The active sensing and optimal control of abnormal events are the core issues to ensure the reliable operation of discrete manufacturing system.For a discrete manufacturing system,many uncertain factors exist in the production system,the degradation patterns of production performance are hard to describe,and the cloud-based data fusion methods consume a long time,as a result,the detection of abnormal events often delays and the optimal decisions are hard to find.To meet these requirements,it is provided a data-driven production performance degradation mechanism and predictive control method for discrete manufacturing system.Firstly,the industrial Internet of Things and Cyber Physical System technologies are combined to establish a cloud-edge cooperation environment and extract the key production performance information.Secondly,the degradation mechanism of production performance is modelled based on the fusion of spatial-temporal manufacturing data.Thirdly,the operational knowledge and predicted performance of manufacturing system will be combined to present an exception early-warning and proactively optimal control method.Then,the capacity of cloud-edge cooperation,early-warning of production exceptions and predictive and optimal control of smart discrete manufacturing system can be achieved.The proposed strategy,method and model provide the important support and technical reference for the predictive control of next generation smart factory.
smart manufacturingself-adaptive optimizationdiscrete manufacturing systemperformance degradationpredictive control
万方数据
维普
