Study On The Fiber Structure Monitoring System And Its Self-diagnostic / Self-healingbased On Multi-agent Collaboration

In the field of aerospace, bridge and tunnel, their structure form and size scale become increasingly large and complicated,the number of sensors for structural health monitoring is increasing,and conventional structural health monitoring system can hardly meet the requirements of precision and reliability. As a new research result in artificial intelligence field, the main idea of multi-agent coordination technology is to accomplish the complex problem by multiple agents interacting together more than by a single agent, whose capacity and knowledge are limited. In this paper, from the aspects of theory and experiment, the multi-agent technology is applied to the field of structural health monitoring to build the fiber intelligent health monitoring system. With the introduction of optical switch and graph theory, the self-healing of optical fiber sensing network was studied when link failures occurred. Using a large aviation aluminum board as the monitoring object, the reliability and corresponding load identification accuracy of the monitoring system are studied by both simulation and experiment. The main content of the paper includes the following sections.1) The concept, model and system structure of agent and multi-agent technology are described, and the multi-agent communication and collaboration are mainly analyzed. On this basis, the frame structure of distributed optical fiber intelligent monitoring system is built based on the multi-agent collaboration.2) To improve the large optical fiber sensing network reliability in the practical application environment, optical switch is introduced in the multi-agent optical fiber sensing network. The optical fiber sensing network topology is designed based on multilateral sensor cells. At the same time, the self-healing and scalability of those sensor networks are studied.3) To repair the link failures which occurred in the optical fiber sensing network, a mesh fiber optic sensor network based on optical switch is proposed, applying related theory of graph theory to it and planning the optical switch switching to realize the self-healing of failure FBG sensor’ signals. And the optical fiber sensing network self-healing effect has been checked by the simulation and experiment. The results show that under the light switch and agent coordination, the load identification accuracy and the reliability of the optical fiber sensing network can be effectively improved.4) To improve impact location accuracy in border region of large structure, the corresponding agent network collaboration model in the two agents regional boundary and four agents regional boundary point are established. By impact load monitoring experiments, the collaboration model can effectively improve the accuracy of impact load location, and the reliability of the whole monitoring system is improved.5) To improve the accuracy of the large structural impact load location, impact load location monitoring method which is based on two different impact load location methods, energy amplitude ratio and the largest inner product match, is proposed. Through the multi-agent blackboard collaboration technology effective fusion of the superiority of two impact load location methods, the experimental results show that the method can improve the positioning precision of impact load of the structure.