Error management in ATLAS TDAQ : an intelligent systems approach
Sloper, John Erik
This thesis is concerned with the use of intelligent system techniques (IST) within\ud a large distributed software system, specifically the ATLAS TDAQ system which\ud has been developed and is currently in use at the European Laboratory for Particle\ud Physics(CERN). The overall aim is to investigate and evaluate a range of ITS\ud techniques in order to improve the error management system (EMS) currently used\ud within the TDAQ system via error detection and classification. The thesis work\ud will provide a reference for future research and development of such methods in the\ud TDAQ system.\ud The thesis begins by describing the TDAQ system and the existing EMS, with a\ud focus on the underlying expert system approach, in order to identify areas where\ud improvements can be made using IST techniques. It then discusses measures of\ud evaluating error detection and classification techniques and the factors specific to\ud the TDAQ system.\ud Error conditions are then simulated in a controlled manner using an experimental\ud setup and datasets were gathered from two different sources. Analysis and processing\ud of the datasets using statistical and ITS techniques shows that clusters exists in\ud the data corresponding to the different simulated errors.\ud Different ITS techniques are applied to the gathered datasets in order to realise an\ud error detection model. These techniques include Artificial Neural Networks (ANNs),\ud Support Vector Machines (SVMs) and Cartesian Genetic Programming (CGP) and\ud a comparison of the respective advantages and disadvantages is made.\ud The principle conclusions from this work are that IST can be successfully used to\ud detect errors in the ATLAS TDAQ system and thus can provide a tool to improve\ud the overall error management system. It is of particular importance that the IST can\ud be used without having a detailed knowledge of the system, as the ATLAS TDAQ\ud is too complex for a single person to have complete understanding of. The results\ud of this research will benefit researchers developing and evaluating IST techniques in\ud similar large scale distributed systems.
views in local repository
downloads in local repository
The information is available from the following content providers: