Wind Vibration Reliability Analysis Of Glass Curtain Wall Based On Swarm Intelligence Algorithm

Along with the advance of science and technology and the rapid development ofnational economy, the high and large span glass curtain wall have been used more andmore. The wind-induced vibration is becoming more and more significant. Ultra-high,large-span glass curtain wall suffered severe damages, caused the loss of people’s lifeand property. So the wind vibration dynamic reliability research has importanttheoretical and practical significance.Difference from analysis methods of component reliability, the support system isso complex and multiple that its performance function is not easy to determine, thecommon reliability methods cannot consider the reliability of point supported glasscurtain wall under wind load. This dissertation uses swarm intelligence algorithm tosolve the structure reliability, provides a new way of thinking for structural reliabilityanalysis.The main work of this dissertation is as follows：(1) The status and future of the point-supported glass curtain wall were analyzed,and its existent problems were derived. The theory of reliability was investigated andthe relative merits of the common reliability analysis approaches were summarized.At the same time, the shortcomings of the traditional structural reliability calculationmethods were pointed out.(2) Respectively, the ant colony algorithm and particle swarm algorithm wereimproved. Through the introduction of information entropy, the adaptive regulation ofant colony algorithm was achieved. The strategies of the inertia weight, the smoothfunction, the threshold function, realize the fast convergence of particle swarmalgorithm on solving the constrained optimization problems．(3) Based on random vibration theory, the response of the structure underrandom wind load and the formula of wind vibration factor were developed. Theveracity of this formula was proved by the results of numerical simulation. Then, thesupporting structures which are fixed-supported and not on the ground were studied.Based on the results of former studies, differently-supported revise andover-ground-based revise are developed for wider application. This formula can beconveniently applied in engineering practice for its comparability with the formula in the load code and it also can be regarded as a reference for code revision.(4) The ant colony algorithm of heuristic information was used to find out thekey structural failure paths, and then the particle swarm optimization algorithm wasemployed to seek the optimal structural failure paths. After establishing of themathematical model, the structural reliability index was calculated. Finally the abovemethods were used in the analysis of the wind vibration reliability of a space trusssupporting system of point supported glass curtain wall.Structural reliability theory is the basis of structure design and specification. Inorder to overcome the shortcomings of the traditional calculation methods, the hybridalgorithms of improved ant colony and particle swarm algorithm were applied tostructural reliability analysis. It achieves better effect on the precision and efficiencyand the study has important significance for improving the level of structure reliabilityanalysis of our country.