A wireless distributed sensor network with low-cost vibration sensors for structural health monitoring

The amount of infrastructure assets such as bridges, process installations and off-shore wind turbines is increasing, accompanied by a rising need for advanced maintenance strategies to reduce inspection and maintenance costs. These maintenance strategies require reliable maintenance status information. Vibration sensing can supply added value for structural integrity monitoring. Large structures, however, require the use of a large amount of sensors which increases the amount of collected data. This puts its demands on data processing and cabling in wired applications with centralized processing. In this paper, a wireless solution is described by means of outlining the development of a distributed sensor network using low-cost wireless vibration sensors. The current network is designed to perform operational modal analysis by applying a relatively simple peak picking algorithm. In this way the eigenfrequencies and mode shapes of a structure can be monitored, which give input to methods for damage detection and characterization. As wireless sensors have a limited communication bandwidth and concurrent data transmissions will flood radio traffic, decentralized data processing is performed to reduce the required amount of communication. Attention will be given to the processing architecture and data flow as well as methods applied to reduce energy consumption. Preliminary results of tests on a scale model of a wind turbine will be presented to assess the quality of the approach. Further, a connection is made with an analysis and optimization tool called Dynamic Adaptive multi sensor networks Architectures (DynAA). With this tool, among others, an operation lifetime analysis can be made with respect to energy-limited, distributed sensor networks. Also, the tool is utilized for optimizing communication parameters in order to extend the maximum operation time of the network.