Engineered system diagnostics have been researched over the years with many successful results. From transportation systems to office technologies, many have been equipped with self-diagnostic capabilities and are called Smart Machines. In spite of these advances, current diagnostic systems are driven by direct sensory information without much concern for patterns of the system behavior or features associated with them. For large-scale systems with complex dynamics, global as well as local diagnostics become of great importance, where sensory information is used as input for the local diagnostics, and patterns of behavior or features are utilized for global diagnostics.
The main objective of this research is to develop a bio-inspired data/information architecture for feature-based global diagnostics of a large-scale system. In order to accomplish this goal, information from local diagnostic systems is integrated with physical and engineering principles (e.g. conservation of momentum) to create a feature-based neuro-symbolic network. This network is very similar to a neural network, except that it is based on a physical equation, and it uses features instead of raw data. Results from this network identify patterns of behavior that display whether the system has any faulty states. The architecture of this network includes two feature-based neuro-symbolic networks, one for the x-axis and one for the z-axis. Each of these networks has four inputs and four outputs, and includes one hidden layer.
In order to verify performance of this feature-based diagnostic system, a working aircraft model has been equipped with pressure sensors on its wing surface, and data has been taken during flight. By using computational fluid dynamic analysis, pressure contours along the wing have been developed. These pressure patterns have been used as input for our developed feature-based neuro-symbolic networks to predict the state of the model aircraft in real-time. Based on these research results, we have developed a neuro-symbolic diagnostic technique which can be applied to fault detection of large-scale systems, using local diagnostic information.
","abstract_html":"<p>Engineered system diagnostics have been researched over the years with many successful results. From transportation systems to office technologies, many have been equipped with self-diagnostic capabilities and are called Smart Machines. In spite of these advances, current diagnostic systems are driven by direct sensory information without much concern for patterns of the system behavior or features associated with them. For large-scale systems with complex dynamics, global as well as local diagnostics become of great importance, where sensory information is used as input for the local diagnostics, and patterns of behavior or features are utilized for global diagnostics.</p> <p>The main objective of this research is to develop a bio-inspired data/information architecture for feature-based global diagnostics of a large-scale system. In order to accomplish this goal, information from local diagnostic systems is integrated with physical and engineering principles (e.g. conservation of momentum) to create a feature-based neuro-symbolic network. This network is very similar to a neural network, except that it is based on a physical equation, and it uses features instead of raw data. Results from this network identify patterns of behavior that display whether the system has any faulty states. The architecture of this network includes two feature-based neuro-symbolic networks, one for the x-axis and one for the z-axis. Each of these networks has four inputs and four outputs, and includes one hidden layer.</p> <p>In order to verify performance of this feature-based diagnostic system, a working aircraft model has been equipped with pressure sensors on its wing surface, and data has been taken during flight. By using computational fluid dynamic analysis, pressure contours along the wing have been developed. These pressure patterns have been used as input for our developed feature-based neuro-symbolic networks to predict the state of the model aircraft in real-time. Based on these research results, we have developed a neuro-symbolic diagnostic technique which can be applied to fault detection of large-scale systems, using local diagnostic information.</p>","abstract_has_math":false,"creators":["Schantz, Tracy Lynn"],"institution":null,"degree_name":"M.S.","degree_level":"Masters Thesis","degree_discipline":null,"degree_department":null,"school":null,"contributors":["Rahmat Shoureshi, Ph.D.","David Gao","Corinne Lengsfeld","Yun Bo Yi"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2012,"date_issued":"2012-01-01T08:00:00Z","date_published":"2012-01-01T08:00:00Z","updated_at":"2026-07-24T02:03:35Z","subjects":["Aircraft structural health monitoring","Computational Fluid Dynamics","CFD","Feature-based diagnostics","Global diagnostics","Neural network","Neuro-symbolic network","Computer-Aided Engineering and Design","Engineering","Mechanical Engineering"],"languages":["en"],"rights":["Copyright is held by the author. User is responsible for all copyright compliance.
"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://digitalcommons.du.edu/etd/580","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Rahmat Shoureshi, Ph.D.","David Gao","Corinne Lengsfeld","Yun Bo Yi"]},{"key":"dc:creator","label":"Author","values":["Schantz, Tracy Lynn"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.available","label":"Dc Date Available","values":["2001-01-01T08:00:00Z"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Masters Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M.S."]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Aircraft structural health monitoring","Computational Fluid Dynamics","CFD","Feature-based diagnostics","Global diagnostics","Neural network","Neuro-symbolic network","Computer-Aided Engineering and Design","Engineering","Mechanical Engineering"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright is held by the author. User is responsible for all copyright compliance.
"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://digitalcommons.du.edu/etd/580"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Engineered system diagnostics have been researched over the years with many successful results. From transportation systems to office technologies, many have been equipped with self-diagnostic capabilities and are called Smart Machines. In spite of these advances, current diagnostic systems are driven by direct sensory information without much concern for patterns of the system behavior or features associated with them. For large-scale systems with complex dynamics, global as well as local diagnostics become of great importance, where sensory information is used as input for the local diagnostics, and patterns of behavior or features are utilized for global diagnostics.
The main objective of this research is to develop a bio-inspired data/information architecture for feature-based global diagnostics of a large-scale system. In order to accomplish this goal, information from local diagnostic systems is integrated with physical and engineering principles (e.g. conservation of momentum) to create a feature-based neuro-symbolic network. This network is very similar to a neural network, except that it is based on a physical equation, and it uses features instead of raw data. Results from this network identify patterns of behavior that display whether the system has any faulty states. The architecture of this network includes two feature-based neuro-symbolic networks, one for the x-axis and one for the z-axis. Each of these networks has four inputs and four outputs, and includes one hidden layer.
In order to verify performance of this feature-based diagnostic system, a working aircraft model has been equipped with pressure sensors on its wing surface, and data has been taken during flight. By using computational fluid dynamic analysis, pressure contours along the wing have been developed. These pressure patterns have been used as input for our developed feature-based neuro-symbolic networks to predict the state of the model aircraft in real-time. Based on these research results, we have developed a neuro-symbolic diagnostic technique which can be applied to fault detection of large-scale systems, using local diagnostic information.
"]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Feature-Based Neuro-Symbolic Networks for Global Diagnostics"]}]}],"canonical_facts":{"dc:contributor":["Rahmat Shoureshi, Ph.D.","David Gao","Corinne Lengsfeld","Yun Bo Yi"],"dc:creator":["Schantz, Tracy Lynn"],"dc:date.available":["2001-01-01T08:00:00Z"],"dc:description.abstract":["Engineered system diagnostics have been researched over the years with many successful results. From transportation systems to office technologies, many have been equipped with self-diagnostic capabilities and are called Smart Machines. In spite of these advances, current diagnostic systems are driven by direct sensory information without much concern for patterns of the system behavior or features associated with them. For large-scale systems with complex dynamics, global as well as local diagnostics become of great importance, where sensory information is used as input for the local diagnostics, and patterns of behavior or features are utilized for global diagnostics.
The main objective of this research is to develop a bio-inspired data/information architecture for feature-based global diagnostics of a large-scale system. In order to accomplish this goal, information from local diagnostic systems is integrated with physical and engineering principles (e.g. conservation of momentum) to create a feature-based neuro-symbolic network. This network is very similar to a neural network, except that it is based on a physical equation, and it uses features instead of raw data. Results from this network identify patterns of behavior that display whether the system has any faulty states. The architecture of this network includes two feature-based neuro-symbolic networks, one for the x-axis and one for the z-axis. Each of these networks has four inputs and four outputs, and includes one hidden layer.
In order to verify performance of this feature-based diagnostic system, a working aircraft model has been equipped with pressure sensors on its wing surface, and data has been taken during flight. By using computational fluid dynamic analysis, pressure contours along the wing have been developed. These pressure patterns have been used as input for our developed feature-based neuro-symbolic networks to predict the state of the model aircraft in real-time. Based on these research results, we have developed a neuro-symbolic diagnostic technique which can be applied to fault detection of large-scale systems, using local diagnostic information.
"],"dc:format":["application/pdf"],"dc:identifier":["https://digitalcommons.du.edu/etd/580"],"dc:language":["en"],"dc:rights":["Copyright is held by the author. User is responsible for all copyright compliance.
"],"dc:subject":["Aircraft structural health monitoring","Computational Fluid Dynamics","CFD","Feature-based diagnostics","Global diagnostics","Neural network","Neuro-symbolic network","Computer-Aided Engineering and Design","Engineering","Mechanical Engineering"],"dc:title":["Feature-Based Neuro-Symbolic Networks for Global Diagnostics"],"thesis:degree_level":["Masters Thesis"],"thesis:degree_name":["M.S."]},"updated_at":"2026-07-24T02:03:35Z"}