Intelligent Software Associates and ExperTech will develop algorithms and tools enabling the rapid construction and cost effective maintenance of embeddable software to reliably detect rocket engine sensor failures during critical launch operations. Over the last four years, NASA and its contractors have developed an advanced approach to propulsion system sensor data validation. Several significant research problems remain to be solved before this approach can be extended to a flight-certified system. We address these problems through the development of an innovative set of software tools which: 1 ) extend the operation of the validation algorithm into all operating states of the engine; 2) provide analyses of the validation system configuration to indicate areas of weakness and assess the overall reliability of the software-based validation; 3) provide automatic tuning of the hundreds of parameters in the validation system; and 4) provide a configuration management utility to quickly adapt the validation system for changes in engine configuration. This research enables a real-time sensor validation capability to reduce operations costs and improve operating safety for NASA propulsion systems by preventing erroneous engine shut-downs and launch aborts based on data from anomalous or failed sensors. The Phase I SBIR effort is fundamental to the commercialization of this technology. Phase I provides essential proof-of-principle testing for a large sensor network in a highly dynamic process control environment. To date, development of models and networks has been performed manually. Automation of this process in Phase I will make commercial sensor validation systems affordable and maintainable. Sensor validation technology is commercially viable for process or vehicle control applications where: 1) time-critical, closed-loop system control and safety monitoring is necessary; or 2) unexpected process interruptions due to sensor failures are un-economical. Related work, funded by the Electric Power Research Institute (EPRI), suggests utility and chemical process control systems offer an unfilled niche for sensor validation technology. We will seek support from EPRI and one or more utility companies for a Phase II demonstration and Phase III production.
Keywords: Phase_I, NASA, Abstract, FY94
