针对结构健康监测的无线传感网校准问题研究

Wireless sensor network (WSN) is one of key technologies in Internet of Things (IOT), which can be applied in many areas, such as structral monitoring, environmental monitoring, and smart transportation et al. Calibration is one of the most critical problems in WSN. The WSN-based structural monitoring system may encouter the obstacles, such as signal discontinuity, sparse deployed sensors, time-variant calibration parameters and nonlinear calibration functions. Against such obstacles, the calibration models will be proposed. This proposal will also study the calibration condition and sensor nodes optimization techniques. Finally, the WSN calibration platform will be constructed.

无线传感网是物联网的关键技术之一，可应用于结构监测、环境监测、智能交通等诸多领域。当前，校准问题是无线传感网大规模应用面临的关键问题之一。本项目拟针对结构健康监测应用中无线传感网所具有的信号不连续、传感器稀疏、时变校准系数、非线性校准函数等特点，提出针对性的校准模型，研究相应的可校准条件和节点优化策略，并构建无线传感网校准平台。

无线传感网技术因其灵活部署、维护方便、适合大规模应用等特点被广泛应用于结构监测、环境监测等领域。但传感器在长期部署的监测应用中易受老化、环境等多方面因素影响而逐渐出现漂移，影响测量准确性。本项目针对无线传感网的校准问题，面向结构健康监测应用，研究大规模自组织组网策略、原位校准方法和测试平台。实现了大量监测数据的同步采集与可靠传输，达到世界领先的10µs同步精度; 提出了基于稀疏贝叶斯学习的传感器原位盲校准算法，可校准传感网中30%的漂移传感器；搭建了测试与仿真平台，验证了盲校准算法。