This paper proposes an artificial neural network(ANN)to determine the optimal foundation isolation scheme by using seismic metamaterial(SM)for building structures.A non-linear relationship was established between SM p...This paper proposes an artificial neural network(ANN)to determine the optimal foundation isolation scheme by using seismic metamaterial(SM)for building structures.A non-linear relationship was established between SM parameters,site parameters,dynamic structural characteristics,and average isolation rate.A database was constructed by employing the fast multipole boundary element method with high efficiency.The problem of the optimal isolation scheme selection was converted to a series of average isolation rate solutions based on the ANN model.The SM-related parameters corresponding to the maximum isolation rate was obtained.The results indicate that ANN has high accuracy in predicting the average isolation rate,with a coefficient of determination of 0.98.The parameter sensitivity measure was similar,indicating the complexity of determining SM layout conditions.The seismic isolation performance of SM was found to be significant for structures with different fundamental frequencies.At the same time,the isolation schemes for structures with the same fundamental frequency varied greatly,with the average isolation rate ranging from 5%to 80%,approximately showing a normal distribution,with a mean value of approximately 55%.展开更多
基金National Natural Science Foundation of China under Grant Nos.52278516 and 52208497Tianjin Science Fund for Distinguished Young Scholars under Grant No.19JCJQJC62900+1 种基金Tianjin“Project+Team”Key Training Project(2020)Special Fund Project for High-Quality Development of Tianjin Manufacturing Industry under Grant No.23ZGCXQY00010。
文摘This paper proposes an artificial neural network(ANN)to determine the optimal foundation isolation scheme by using seismic metamaterial(SM)for building structures.A non-linear relationship was established between SM parameters,site parameters,dynamic structural characteristics,and average isolation rate.A database was constructed by employing the fast multipole boundary element method with high efficiency.The problem of the optimal isolation scheme selection was converted to a series of average isolation rate solutions based on the ANN model.The SM-related parameters corresponding to the maximum isolation rate was obtained.The results indicate that ANN has high accuracy in predicting the average isolation rate,with a coefficient of determination of 0.98.The parameter sensitivity measure was similar,indicating the complexity of determining SM layout conditions.The seismic isolation performance of SM was found to be significant for structures with different fundamental frequencies.At the same time,the isolation schemes for structures with the same fundamental frequency varied greatly,with the average isolation rate ranging from 5%to 80%,approximately showing a normal distribution,with a mean value of approximately 55%.
