| Internet of Things(IoTs)entity discovery plays an important role in the Industrial IoTs(IIoTs)especially with the rapidly increasing and updating of Io T devices in the industrial environment driven by the era of industry 4.0 and intelligent manufacturing.However,large numbers of non-smart devices are required in the industrial environment causing the challenge on IoTs entity discovery.Different from the smart device,the nonsmart device with limited computation and communication ability is hard to be discovered and recognized actively by the traditional IoTs platforms.And existing research on datadriven device discovery only achieves better performance in the dataset,but with the increase and update of devices,the model effect decreases gradually,and the model needs to be trained again.Aiming at the challenge,this work proposes a novel IoTs entity discovery middleware for non-smart sensor discovery in the industrial environment.The proposed middleware combines both device knowledge graph and device data value to build an IoTs entity discovery and recognition model.The main contributions of the thesis are the follows:(1)A non-smart device knowledge graph that not only provides high-level information(such as type,brand,attributes,input,output,etc.)of devices but also contains low-level information(such as PLC addresses,data points,data ranges,etc.),is proposed.(2)A knowledge-data fused learning network is proposed for the model to identify the data type,function,and other information of the non-smart sensor.In the experimental evaluations,the prototype middleware tests a variety of different non-smart devices and achieves 87.5% recognition accuracy.(3)At last,a prototype middleware with the discovery and recognition model is produced to implement non-smart sensor discovery.In the real-world case studies,the prototype middleware proves the feasibility and effectiveness of non-smart sensor discovery in the industrial environment of rural sewage treatment and copper material production line. |