| Intelligent sensors are investigated as part of an integrated systems approach. This approach has resulted in the development of two types of sensors: Physical Intelligent Sensors (PIS) and Virtual Intelligent Sensors (VIS). The integrated systems approach, as implemented in the PIS, treats the sensor as a complete system with its own sensing hardware (the traditional sensor), A/D converter, processing and storage capabilities, software drivers, self-assessment algorithms and communication protocols. The VIS is the virtual counterpart of the PIS where the algorithms reside on the main computer and is the main focus of this thesis. The VIS is used where it is impossible to place a PIS as well as a test platform for algorithms before downloading them to the PIS. The immediate application is the monitoring of rocket test stands at NASA's Stennis Space Center, but the technology should be generally applicable to the Integrated Systems Health Monitoring (ISHM) vision. This thesis outlines progress made in the development of intelligent sensors by describing the work done to date on Virtual Intelligent Sensors (VIS) with implementation in the LabVIEW environments. The VIS has been designed to mirror the operations of the PIS, and uses existing anomaly detection routines developed at NASA. This thesis also focuses on smart sensors, and shows the working of an enhanced version of a previously developed intelligent sensor algorithm, DATA-SIMLAMT, called Enhanced DATA-SIMLAMT or EDATA-SIMLAMT. This new version contains additional properties and states for improved data interpretation. The additional properties are based on a signal processing tool known as wavelet analysis. The major advantage provided by adding wavelet analysis is the ability to detect sudden transitions as well as get a handle on the frequency content using a much smaller data set then that required by the traditional Fourier transform method. |
