Optimum Design of Structures against Earthquakes using Discrete Wavelet Neural Networks

摘要

Optimum design of structures is achieved against earthquake loads. Genetic algorithm is employed to find the optimal design variables. Time history dynamic analysis is carried out to find the structural responses for each design point. The design constraints are evaluated at all the time intervals of the earthquake. To reduce the computational time of optimisation, a fast wavelet transform (FWT) is used. By FWT the time intervals of the record is reduced. In addition, by a wavelet neural network (WNN), the dynamic responses of the structure under consideration are approximated. By such approximation, the dynamic analysis of the structure is not necessary during the optimisation process. WNN is employed as a general approximation tool for the time history dynamic analysis. To reconstruct the actual responses of the structure from the responses of the structure against FWT points (output of WNN), a reverse wavelet transform (RWT) is employed. A number of structures are designed for optimal weight and the results are compared with exact dynamic analysis.