- Related Research Areas
-
Diagnostics
- Project Description
- The Third International Diagnostic Competition (DXC'11) was hosted at the 22nd International Workshop on Principles of Diagnosis (DX-11) in Murnau, Germany, October 2011. The objectives of the competition series are to accelerate research in theories, principles, and computational techniques for monitoring and diagnosis of complex systems, to e...Morencourage the development of diagnostic software platforms, and to provide a forum that can be utilized by algorithm developers to test and validate their technologies on real-world physical systems.
The First International Diagnostic Competition (DXC'09) DASHlink project page can be accessed here. The Second International Diagnostic Competition (DXC'10) DASHlink project page can be accessed here.
This DASHlink project page focuses on the Industrial Track of the competition. For more information on the Synthetic and Software tracks, click here
Using the DXC'11 Data and Information
The competition has concluded but we provide DXC'11 data and information on this project page for researchers who would like to evaluate their diagnostic algorithms (DAs) using data sets from a physical system and compare the performance against other DAs. Here's an overview of the steps involved.
1.) Knowledge acquisition. Refer to the resources page for detailed information about the Industrial Track diagnostic problems (note that the problem formulation is the same as DXC'10), sample data, and system description catalogs and schematics.
2.) Develop diagnostic algorithms/models. Using the documentation and sample data, DA developers build diagnostic models and strategies to diagnose injected faults.
3.) Interface DA with DXC framework. The DA uses framework API calls to read sensor data and commands from the scenario files and provides diagnoses and recommended recovery actions.
4.) Evaluate DA performance on sample scenarios. Using evaluator software provided with the framework, DA developers assess the performance on sample scenarios and make improvements as necessary.
5.) Evaluate DA performance on competition scenarios. Now, try out the DA on the competition data and see how it compares to results from DXC'11.
6.) Refer to the papers and slides for additional information....Less
- Project Administrator(s):
-
SCOTT POLL
Members
×
