{"@type": "dcat:Dataset", "accessLevel": "public", "accrualPeriodicity": "irregular", "bureauCode": ["026:00"], "contactPoint": {"@type": "vcard:Contact", "fn": "SCOTT POLL", "hasEmail": "mailto:scott.d.poll@nasa.gov"}, "description": "A variety of rule-based, model-based and datadriven\r\ntechniques have been proposed for detection\r\nand isolation of faults in physical systems.\r\nHowever, there have been few efforts to comparatively\r\nanalyze the performance of these approaches\r\non the same system under identical conditions.\r\nOne reason for this was the lack of a standard\r\nframework to perform this comparison. In this paper\r\nwe introduce a framework, called DXF, that\r\nprovides a common language to represent the system\r\ndescription, sensor data and the fault diagnosis\r\nresults; a run-time architecture to execute\r\nthe diagnosis algorithms under identical conditions\r\nand collect the diagnosis results; and an evaluation\r\ncomponent that can compute performance\r\nmetrics from the diagnosis results to compare the\r\nalgorithms. We have used DXF to perform an empirical\r\nevaluation of 13 diagnostic algorithms on a\r\nhardware testbed (ADAPT) at NASA Ames Research\r\nCenter and on a set of synthetic circuits\r\ntypically used as benchmarks in the model-based\r\ndiagnosis community. Based on these empirical\r\ndata we analyze the performance of each algorithm\r\nand suggest directions for future development.", "distribution": [{"@type": "dcat:Distribution", "description": "ijPHM_10_002.pdf", "downloadURL": "https://c3.nasa.gov/dashlink/static/media/publication/ijPHM_10_002.pdf", "format": "PDF", "mediaType": "application/pdf", "title": "ijPHM_10_002.pdf"}], "identifier": "DASHLINK_864", "issued": "2013-12-18", "keyword": ["ames", "dashlink", "nasa"], "landingPage": "https://c3.nasa.gov/dashlink/resources/864/", "modified": "2025-03-31", "programCode": ["026:029"], "publisher": {"@type": "org:Organization", "name": "Dashlink"}, "title": "Empirical Evaluation of Diagnostic Algorithm Performance Using a Generic Framework"}