• Projects
  • Active Flow Control
  • Adaptive Reconfigurable Control
  • Diagnostics and Prognostics
    • Adaptive Diagnostics and Prognostics Toolbox (ADAPT)
    • Diesel Engine Diagnostics and Prognostics
    • Sensor and Actuator Fault Detection and Isolation
    • Prognostic Techniques for Mechanical Failure Prediction
    • Structural Health Monitoring
  • Intelligent Guidance and Trajectory Reshaping
  • Real-Time Modeling and Prediction
  • Tools for Healthcare Assessment

Adaptive Diagnostics and Prognostics Toolbox (ADAPT)

Model-based machinery diagnostic and prognostic techniques depend upon high-quality mathematical models of the plant. Modeling uncertainties and errors decrease system sensitivity to faults and decrease the accuracy of failure prognoses. However, the behavior of many physical systems is poorly known and may change slowly over time as the system ages. These changes may be perfectly normal and not indicative of impending failures; however, if a static a priori model is used, modeling errors may increase over time, which can adversely affect health monitoring system performance. Moreover, manufacturing tolerances and environmental impacts may cause variations in unfaulted system dynamics between nominally identical plants. These variations significantly complicate practical, widespread application of model-based diagnostic techniques.

Clearly, one method to address these problems is to employ a model that adapts to system uncertainties and slow changes over time. The risk in using data-driven models that learn online to support model-based diagnostics is that the models may “adapt" to a system failure, thus rendering it undetectable by the diagnostic algorithms. An inherent trade-off exists between accurately tracking normal variations in system dynamics and potentially obscuring slow-onset failures by adapting to failure precursors that would be evident using static models.

In practice, the behavior of many physical systems is complex, encompasses a large operating regime, and is subject to external disturbances and noise. Exhaustive data collection to develop highly accurate a priori models of unfaulted system behavior may be costly and impractical in many applications. It would be highly valuable to utilize an automated, readily portable set of integrated adaptive diagnostic and control algorithms for a large class of dynamic systems. 

An adaptive diagnostic and prognostic system has three main components: (1) online system identification / parameter estimation; (2) adaptive modeling/filtering/observation, and; (3) diagnostic and prognostic algorithms and logic. ADAPT is a Matlab/Simulink Toolbox constructed of modular components (Simulink blocks) that each contain software that executes a specific algorithm from one of the aforementioned categories. The user is able to mix-and-match algorithms from each functional category to construct an integrated adaptive diagnostic and prognostic system.

Recent Applications and Customers 

Barron Associates, with its research partners at the University of Virginia, has recently conducted extensive research in the areas of adaptive diagnostic and prognostic technologies – primarily funded by NASA Langley Research Center as part of NASA’s Aviation Safety Program.  ADAPT is widely applicable to many linear and nonlinear dynamic systems subject to actuator, sensor, and plant faults and failures.  Please contact Barron Associates for information on obtaining the ADAPT Toolbox. 

Contact Info

Jason Burkholder:  burkholder@barron-associates.com