Action Filename Description Size Access License Resource Version
Show more files...


Biomimetic structures are structures that demonstrate increased functionality through mimicking qualities of biological organisms. Self-repair and adaptation mechanisms are examples of biological qualities that can be adapted in structural engineering. Over the last decades, great strides have been made in advancing theory and practice of active structural control. However, little scientific progress has been made on biomimetic structures. Advances in sensor, actuator, and microprocessor technologies provide increasing possibilities for implementing active control systems in the built environment. Intelligent control methodologies such as self-diagnosis, self-repair and learning could be integrated into structural systems to provide innovative solutions. The general goal of this thesis is to study biomimetic characteristics of an active and deployable tensegrity bridge. Building on previous research carried out at EPFL, this thesis proposal includes the following objectives: 1) design an active control system in order to ensure damage tolerance of a deployable tensegrity pedestrian bridge; 2) extend existing strategies for self-diagnosis of the deployable tensegrity bridge to avoid ambiguous results; 3) extend existing strategies in order to achieve a more robust self-repair scheme; 4) develop algorithms that allow the active control system to learn efficiently using case-based reasoning; 5) validate the methodologies developed with experiments on a near full-scale (1/3) model. A literature survey of biomimetics, structural control, tensegrity structures, deployable structures, deployable tensegrity structures, active tensegrity structures, case-based reasoning, system identification, and multi-objective search has identified that these objectives are original. Results obtained from the preliminary studies demonstrate the potential of this research strategy. A research plan containing 19 subtasks that will be completed by the end of April 2012 leaves sufficient buffer time before the official end of this Ph.D. research on September 30, 2012.