Dynamic wireless sensor network-based structural health monitoring system for retractable roof structure

Retractable roof structure (RRS) works optimally under varying weather, but is prone to damage due to its complexity, requiring monitoring. Given RRS's large scale, wireless sensor networks (WSNs) are preferred, with dynamic WSNs better adapted to its motion than static ones. A tree- type network topology was designed for dynamic WSN, enabling fast self-routing. It is divided into subnets, each with a relay node as the parent and subordinate relay and sensor nodes as children. Network address reassignment method was proposed to sustain WSN communication during motion. This method involves operations of subnet affiliation change and relay node replacement, the former adjusts the hierarchy between relay nodes across different subnets while preserving parent-child relation within each subnet, and the latter swaps the roles of relay and sensor nodes within one subnet to form a new subnet. Basic communication module with processing, communication, and power management units was used to connect nodes. Dynamic WSN-based structural health monitoring (SHM) system with 372 sensors was developed for the Textile City Sports Center Gymnasium. Mean errors between simulated and measured results were 10.4 % for displacement, 13.6 % for stress during construction, and 13.3 % for tension during service, indicating good agreement. This suggests that the SHM system operated well in both phases. Maximal stress and displacement after unloading were 215.9 MPa and 549 mm, below allowable limits. Monitoring results revealed temperature field non-uniformity with a 9 degrees C maximal difference, and strong temperature-stress correlation, with Pearson coefficients of 0.971, 0.973, and 0.955 for tie rod, chord, and web. RRS movement significantly impacted midspan tie rods in long main trusses, with a mean tension change exceeding 1000 kN. The research methods and conclusions offer valuable references for developing and applying RRS's SHM system.