dc.contributor.author |
El Itani, Rafic Mohammad, |
dc.date.accessioned |
2017-12-11T16:30:52Z |
dc.date.available |
2017-12-11T16:30:52Z |
dc.date.issued |
2017 |
dc.date.submitted |
2017 |
dc.identifier.other |
b19133339 |
dc.identifier.uri |
http://hdl.handle.net/10938/20987 |
dc.description |
Thesis. M.E.M. American University of Beirut. Department of Industrial Engineering and Management, 2017. ET:6539 |
dc.description |
Advisor : Dr. Hussein Tarhini, Assistant Professor, Industrial Engineering and Management ; Committee members : Dr. Bacel Maddah, Associate Professor, Industrial Engineering and Management ; Dr. Samir Mustapha, Assistant Professor, Mechanical Engineering. |
dc.description |
Includes bibliographical references (leaves 64-65) |
dc.description.abstract |
Sensor placement for physical fault detection on convex plane surfaces is a quick spreading technology that has been used in evolving industrial firms for structural health monitoring purposes. A set of sensors is allocated on the surface understudy and ultrasonic guided waves are excited between sensor pairs to detect and allocate possible damage. The cost of installation and maintenance of such structural health monitoring systems has been reported to grow exponentially as the size of the structure increases. Hence, several techniques for optimizing sensor networks have been presented in literature to improve coverage and fault detection while reducing the number of sensors needed and hence reducing cost. In this paper, a demonstration of the preceding sensor network optimization approaches is presented and an advanced geometrical optimization approach for fault detection and sensor placement is proposed. The approach is formulated as a Mixed Integer Non-Linear program (MINLP) with user defined parameters to simulate actual geometrical conditions of the surface understudy and sensor coverage characteristics. The model is tested in several real case studies with different scenarios of coverage levels for both symmetrical and optimized sensor arrays and assures the efficiency and strong performance of the aforementioned approach. Data fusion is also carried out for the optimal sensor locations determined by the experimentation scenarios and the results confirm the paper findings. |
dc.format.extent |
1 online resource (ix, 65 leaves) : color illustrations |
dc.language.iso |
eng |
dc.relation.ispartof |
Theses, Dissertations, and Projects |
dc.subject.classification |
ET:006539 |
dc.subject.lcsh |
Sensor networks. |
dc.subject.lcsh |
Genetic algorithms. |
dc.subject.lcsh |
Simulated annealing (Mathematics) |
dc.subject.lcsh |
Structural health monitoring. |
dc.subject.lcsh |
Combinatorial optimization. |
dc.title |
Sensor network optimization for structural health monitoring - |
dc.type |
Thesis |
dc.contributor.department |
Faculty of Engineering and Architecture. |
dc.contributor.department |
Department of Industrial Engineering and Management, |
dc.contributor.institution |
American University of Beirut. |