Abstract Since the wavelet neural network has the defects of slow convergence rate and proneness of local opti? mum, while the artificial bee colony algorithm is capable of local and global search, an optimization of wavelet neural network (WNN) was conducted using artificial bee colony (ABC) algorithm to form artificial bee colony-wavelet neural network (ABC-WNN) and it was applied to prediction of settlement of metro tunnels. Taking the Shenzhen metro line 10 for example, a comparison and analysis were carried out regarding the predicted results and that of BP neural network (BPNN) and WNN. The results show that the prediction result of ABC-WNN is more accurate and stable than that of other two models.
Abstract：
Since the wavelet neural network has the defects of slow convergence rate and proneness of local opti? mum, while the artificial bee colony algorithm is capable of local and global search, an optimization of wavelet neural network (WNN) was conducted using artificial bee colony (ABC) algorithm to form artificial bee colony-wavelet neural network (ABC-WNN) and it was applied to prediction of settlement of metro tunnels. Taking the Shenzhen metro line 10 for example, a comparison and analysis were carried out regarding the predicted results and that of BP neural network (BPNN) and WNN. The results show that the prediction result of ABC-WNN is more accurate and stable than that of other two models.
CHEN Youzhou1 REN Tao2 DENG Peng2 WANG Bin3
.Prediction of Tunnel Settlements by Optimized Wavelet Neural Network Based on ABC[J] MODERN TUNNELLING TECHNOLOGY, 2019,V56(4): 56-61
2019, Vol. 56 