Intelligent agent based, autonomous real-time structural health monitoring system

Active structural health monitoring enables real time decisions about the integrity of structural system through visualization of the extent of damage in real-time. The work presented here explains the development of an autonomous real-time structural health monitoring system composed of three pieces: the sensor hardware, the autonomous task planner, and the health monitoring system. The sensor hardware is designed to process the sensor data in the sensor nodes and transmit the processed data wirelessly to a remote location. Attention has been paid to size, weight, power consumption, range of inputs, and transmission distance. Processing data at the sensors prior to transmission aids in scalability in the size of the sensor network by controlling the amount of data that needs to be transmitted around the sensor array. Processing data at the sensor using a methodology allowing reconfigurable algorithms is a new concept in data analysis. The health monitoring system is a collection of algorithms that process sampled data in either the time or frequency domain to produce a measure of the integrity of the structure. Each domain has its advantages so algorithms for both have been developed. The autonomous task planner is based on using intelligent agent theory to add autonomous actions in both task planning and health monitoring. Agents are being used to route communications, detect anomalies in the data and form sub groups to investigate these anomalies. The sensor node hardware, the intelligent agents and the structural health monitoring algorithms have been combined to demonstrate the operation of the integrated system. The integrated system monitors the structural system in real-time from sending information about system integrity to a remote location.