A sensor network event fusion and decision method based on node reliability

Abstract

The invention discloses a sensor network event fusion and decision method based on node reliability. The sensor network is used to judge and decide the structural engineering operation conditions, including event classification and event decision fusion.. In the phase of event classification, the sensor network detects and classifies abnormal events in structural engineering, and reflects the changes of physical state in local area. Considering the reliability difference of nodes, the classification error can be reduced by cooperating with each other. In the event decision fusion stage: abnormal event level changes can reflect the operation state of the structure, and different event classification results are combined to evaluate and make decisions on the structural engineering conditions. Using event fusion model based on depth learning, data enhancement method is used to solve the problem of small training samples, update model parameters dynamically, and assist the evaluation of structural engineering operating conditions.

Description

technical field [0001] The invention relates to a sensor network event fusion and decision-making method based on node reliability, in particular to a sensor network for judging and decision-making of large-scale civil engineering, belonging to the field of sensor network monitoring. Background technique [0002] With the development of embedded sensing technology, large civil and structural projects (bridges, signal towers, dams, tunnels, etc.) are buried with a large number of sensors during the construction period and interconnected to form a sensor network to monitor physical quantities such as deformation, stress, strain, and displacement. The sensor network perceives changes in the physical quantities of structural engineering, detects the occurrence of abnormal events and determines the event category, and then fuses the event classification results based on professional models to judge the operating status of structural engineering. [0003] The evaluation of the ope...

AI Technical Summary

Problems solved by technology

[0006] Purpose of the invention: In order to improve the decision-making accuracy and real-time performance of structural engineering operating conditions, the present invention provides a sensor network event fusion and decision-making method based on node reliability, and uses node reliability differences in the event classification stage to improve event classification accuracy In the event decision fusion stage, data enhancement is used to solve the problem of few training samples, and the event fusion model based on deep learning is used to dynamically update model parameters to assist in the evaluation of structural engineering operating conditions

Images