DrMUST - a Data Mining Approach for Anomaly Investigation

Autor:	Bruno Sousa, José-Antonio Martínez-Heras, Jörg Fischer
Rok vydání:	2012
Předmět:	Engineering Spacecraft business.industry Event (computing) computer.software_genre Machine learning Correlation analysis Pattern matching Artificial intelligence Data mining Anomaly (physics) business Time complexity computer Mathematical correlation
Zdroj:	SpaceOps 2012 Conference.
DOI:	10.2514/6.2012-1275109
Popis:	DrMUST is a data mining MUST client that can support flight control engineers in their anomaly investigation tasks. It performs pattern matching and correlation analysis. The pattern matching functionality can be used to find occurrences of a certain behavior (e.g. to know when certain anomaly happened in the past and went unnoticed). The correlation analysis can find which parameters are involved in a certain event of interest (e.g. anomaly). The correlation analysis is based on statistical features and is more robust and efficient than the classical mathematical correlation, allowing to perform correlations in linear time. DrMUST can also be used to characterize how a spacecraft is affected by environmental changes (e.g. solar flares). This work describes the technology behind DrMUST and provides real examples from ESA operated missions.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::bfa4be8328f5ef83caab301c326af790 https://doi.org/10.2514/6.2012-1275109 Zobrazit plný text záznamu