Symbolic encoding of analytic signals for structural monitoring of power systems

This work aims to develop a novel non-destructive technique for real-time condition monitoring to enhance reliability and availability of fossil power systems during steady-state and transient operations. We focused on mechanical fatigue in plant components with the specific objectives of detection and identification of failure precursors, and compression of pertinent damage information to enable transmission over a wireless communication network. A symbolic encoding based technique is employed that utilizes analytic representation of ultrasonic signals for monitoring crack formation and propagation. The signal processing tools are implemented in MATLAB/Simulink environment, for which the direct code generation capability enables the algorithms to be executed on the microprocessor in real-time.