A Sub-Microwatt Self-Powered Fatigue Sensor

提出了一种基于压电传感器与超低功耗浮栅计算电路阵列集成的新型结构疲劳传感器。开发的小型化传感器将实现土木工程结构的连续无电池监测和故障时间预测。通过控制浮栅晶体管中的热电子注入，研究了一种能够计算和存储结构累积疲劳历史的自供电传感系统。该统计数据将与机器学习技术相结合，用于预测剩余疲劳寿命，从而预测结构的失效时间。典型土木工程结构在使用荷载作用下，振动能量转换的可用功率小于10W。这将是首个能够在小于1W的功率水平下进行机载数据存储和计算的自供电疲劳传感器。疲劳传感器将在民用基础设施监测中得到应用（对于难以检查的结构，如桥梁、交通标志和埋在地下的部件，尤其重要）；也可用于其他机械和生物医学疲劳监测。该项目还将促进土木和电气工程之间的紧密合作，并为培养具有跨学科背景的研究生提供理想的媒介。

The development of a novel structural fatigue sensor based on integration of a piezoelectric transducer with an array of ultra-low power floating gate computational circuits is proposed. The developed miniaturized sensor will enable continuous batteryless monitoring and time-to-failure predictions of civil engineering structures. By controlling hot-electron injection in floating gate transistors,a self-powered sensing system capable of computing and storing the cumulative fatigue history experienced by a structure is investigated. This statistics in conjunction with machine learning techniques will be used to predict remaining fatigue life and hence the time-to-failure of the structure. The power available from vibration energy conversion in typical civil engineering structures is less than 10W under service loads. This will be the first self-powered fatigue sensors that is capable of on-board data storage and computation at power levels of less than 1W. The fatigue sensor will have applications in civil infrastructure monitoring (of particular importance for structures that are difficult to inspect such as bridges, traffic signs, and buried components); and also for other mechanical and biomedical fatigue monitoring. This project will also facilitate close collaboration between civil and electrical engineering and provide an ideal medium for training graduate students with cross-disciplinary backgrounds.