Research On Boundary Tracking Of Continuous Objects Based On IIOT For Smart Factory

To ensure the secure and green production requirements of smart factories,two continuous objects tracking algorithms are proposed from the perspectives of delay-sensitivity and predictability,respectively.The algorithms make full use of technologies including temporal-spatial sensing,target trajectory sensing,and mobile edge computing.The performance of each algorithm is analyzed and verified by simulation.The main research findings of this paper are as follows:(1)To meet the delay-sensitive requirement of tasks in continuous object tracking,a continuous object tracking algorithm based on parallel optimization framework is proposed.The algorithm proposes a three-tier computing architecture consisting of sensor nodes,base stations and cloud server,defines the calculation method of each layer,and constructs a joint optimization problem of task offloading and assignment to minimize average delay under the limitation of computing resources.To solve the problem,a parallel optimization framework is proposed.The proposed framework decomposes the optimization problem into three subproblems firstly,and then combines the task offloading and assignment algorithm based on trajectory sensing of continuous object to realize gradual optimization process for decoupling tasks,solving sub-problems,and updating problem solutions.(2)In view of the low-frequency and high-risk characteristics of continuous objects,a two-stage continuous object tracking algorithm based on the state transition model is proposed.The algorithm proposes a transition model of node tracking state which defines the node's tracking states and the specific actions,to realize that the nodes maintain sleep state in idle time to save energy.In order to achieve accurate prediction tracking,a two-stage tracking strategy based on different diffusion stage is proposed,which includes a discrete tracking algorithm in units of nodes and a intercluster cooperative tracking algorithm in units of clusters.A regional wake-up algorithm is proposed to reduce tracking error.Simulation results demonstrate that the tracking algorithm based on a parallel optimization framework reduces the task execution delay in different situations and meets the delay-sensitive requirement of tasks,and shows the two-stage tracking algorithm based on the state transition model has lower energy consumption and communication traffic,while reducing tracking error.