Production total factor analysis system based on hierarchical causal scoring model
In order to comprehensively improve product quality,open up multilevel IT system,opti-mize production planning,and realize tracking and tracing of the whole life cycle of production line,a production total factor analysis system was proposed,which processes real-time asset data through a rule engine to form feature input containing business logic,uses feature input and diagnostic prior knowledge to build a hierarchical causal score graph,and constructs a hierarchical score model at the node level of the graph to obtain inference results from low level to high level of processes,equip-ment,materials,personnel and other nodes.Through layer upon layer reasoning of the directed graph,a score result bound with the production ID number is finally obtained.In this process,the node loca-tion causing the score anomaly can be quickly and accurately located.Map abstract asset information to field production problems,allowing managers to respond quickly.In the field application,the joint analysis of the same batch of different materials can extend the monitoring of the whole production factors such as quality,yield and energy consumption of a single material to the entire production line.The construction of the model is completed through the visual operation page.Due to the excel-lent expansion performance of the model,a rich production analysis model library is built and the pre-cipitation of expert knowledge is realized,which has strong promotion value.
total factor analysis of productionrules enginehierarchical causal score graphinference model
万方数据
维普
