Research On Soft Computing For Optical Fibers Based Intelligent Structures

As an emerging applied technology, damage detection is an important part of structural health monitoring, which ultimate goal is to increase reliability, improve safety and reduce maintenances cost. The controller is the core and the object of controlling,monitoring and evaluating is the smart structure itself. Because the structure stays in the circumstance of time-variability and uncertainty and the structure itself is a complex system, contains fuzzy information expressed as natural language and data information represented as exact number, the controller is demanded to be high degree of intelligence,robustness and learning ability, so the soft computing(SC) is appropriate to smart structure controller. SC has potential to emulate the remarkable human ability to learn and adapt to changes of surrounding, is now an evolving collection of methodologies including Fuzzy system, Neural networks and Genetic Algorithms, designed to exploit tolerance for imprecision, uncertain and partial truth to achieve tractability, robustness at low solution cost. As a novel approach to create computing intelligence SC is coming to draw more attention and we now have recognized that a complex problem such as monitoring for smart structure, only be settled properly by combination of different technologies and methods possibly. The core of intelligence system is soft computing and this paper studies the applications of SC for optic fiber smart structure assessment. Main works and creative conclusions are presented as follow:1.Optic fiber sensors and actuators are united with the composite materials to form the smart structures, therefore it regulate itself to fit the surrounding. The optic fiber function as sensors as well as transmitter,is the most ordinary and promising sensor in smart structure. In this paper, the principle of Fabery-Perot sensor and Bragg grating sensor are presented, the corresponding experimentation and optic fiber smart layer are described.2. Granular Computing (GrC) is a branch of soft computing science, it play an important role in information processing for fuzzy, uncertainty, partial truth and numbered and becomes one of the main study stream of AI. GrC adapts the traditional notion of computing from seeking for absolute optimization to the good enough result. So the information processing is more reasonable,comprehensive,economical and convenient. This paper applies the GrC to the structure assessment and introduces nominal and numerical granule to quantitative analysis nominal and numerical data characteristic. Therefore a more reasonable evaluation is presented upon managing both...