Multi-agent system for tunnel-settlement monitoring: A case study in Shanghai

Abstract

Special soil environment and geographic location expose tunnels in Shanghai to a higher risk of asymmetry change and non-homogeneous settlement that could make a threat to passenger safety and tunnel structure stability. Considering work efficiency and costs, current monitoring methods and frequency are not suitable for tunnel monitoring present stage requirements. To make sure tunnel under condition of economic sustainability, environmental sustainability and engineering safety, new technology and instrument for monitoring should be designed to break through. Besides, asynchrony and retardation existing in data acquisition and analysis makes monitoring task cannot reflect tunnel real-time status. This can make people unable to timely detect the occurrence and risk of tunnel disaster. The purpose of this article is to develop a tunnel settlement monitoring system with several intelligent functions for comprehensive evaluating tunnel healthy, generating seasonable warnings and guaranteeing construction sustainability. To facilitate the expansion of the system and data structure, agent techniques and condition based monitoring standard have been applied to establish the system with 9 action parts according to 3 different geospatial. Moreover, web service has also been used to provide human–computer interface and realize visual function, and meanwhile, all the procedure of the system was completed automatically. Additionally, the system was tested with a task in Shanghai tunnel, and the result shows the monitoring system can describe tunnel conditions, prevent tunnel hazards and assist managers to make appropriate decisions with less people and reduced costs.