A Novel Wireless Sensor Network with Advanced Prognostic Algorithms for Condition Based Maintenance of Critical Power Plant Components
The ability to accurately predict the early stages of failure and the remaining useful life of critical power plant components (such as bearings, turbines, and rotor disks) is critical for affordable system operation and for enhanced system safety. Usually, failures go through a series of transitions from normal operation, to minor degradation, and finally to complete failure. Each stage of degradation generally has unique characteristics that can be identified from sensor outputs. Therefore, an advanced, robust, and reliable algorithm is needed that can perform accurate prognostics. This project will develop an advanced prognostic capability that can be added to a Wireless Sensor Networks (WSN) Â¿ which can provide real-time, continuous data collection Â¿ to monitor critical components in power plants. The hardware will consist of a wireless sensor network with appropriate sensors and data acquisition card, and a portable PC. The portable PC will contain prognostic algorithms based on a Hidden Markov Model that has been shown to be very successful in capturing the transitions between various degraded states. A user-friendly Graphical User Interface (GUI) will display component health status and trends. Commercial Applications and other Benefits as described by the awardee: The advanced WSN, which will combine low cost hardware and innovative prognostic software in a unified framework, should reduce system downtime and maintenance costs. In addition, the technology should find use in any DOE, NASA, military, or commercial application where electromechanical systems include mechanical components (e.g., turbine engines, bearings, pumps, gearboxes, motors, etc).
Small Business Information at Submission:
Signal Processing, Inc.
13619 Valley Oak Circle Rockville, MD 20850
Number of Employees: