Improved RBF neural networks for estimating deformation modulus of hard rock formations
List of Authors
  • Lim Eng Aik , Tan Wee Choon

Keyword
  • Deformation modulus, RBF neural network, Pattern search method, BP neural Network, rock mass.

Abstract
  • Various methods are employed to ascertain the deformation modulus of rock masses, encompassing indoor and outdoor testing, numerical analysis, inverse analysis, and rock mass classification. However, these methods are characterized by notable limitations. Advancements in neural network techniques offer a promising avenue for predicting rock mass parameters through modeling. Utilizing a dataset comprising 88 sets obtained from literature, this study employed rock quality indicators including Rock Quality Designation (RQD), Rock Mass Rating (RMR), and compressional wave velocity (Vp). Leveraging these factors, an enhanced Radial Basis Function (RBF) neural network model was developed, incorporating the pattern search method to predict the deformation modulus of rock masses. To assess the model's accuracy, 17 datasets from the rock masses of the literature taken as test set, and the resultant predictions were compared with those derived from a Backpropagation (BP) neural network model and in-situ data. The findings indicate that the enhanced RBF model exhibits superior performance in predicting the deformation modulus of hard rock masses.

Reference
  • 1. Bataineh, M., & Marler, T. (2017). Neural network for regression problems with reduced training sets. Neural Networks, 95, 1–9.

    2. Grima, M., & Babuska, R. (1999). Fuzzy model for the prediction of unconfined compressive strength of rock samples. International Journal of Rock Mechanics and Mining Sciences, 36(3), 339-349.

    3. Ismkhan, H. (2018). I-k-means++: An iterative clustering algorithm based on an enhanced version of the k-means. Pattern Recognition, 79, 402-413.

    4. Kant, S., & Ansari, I.A. (2016). An improved K means clustering with Atkinson index to classify liver patient dataset. International Journal of System Assurance Engineering and Management, 7, 222-228.

    5. Liu, T., Chen, J., & Sun, Y. (2017). Prediction of deformation modulus of hard rock mass based on RBF neural network optimized by genetic algorithm. Engineering Geology, 226, 36-45.

    6. Liu, A., Chen, X., & Liu, Y. (2019). Prediction of deformation modulus of hard rock mass based on improved RBF neural network optimized by firefly algorithm. Engineering with Computers, 35(3), 1011-1023.

    7. Li, S., Zhang, L., & Wang, X. (2021). Predicting deformation modulus of hard rock mass based on improved extreme learning machine. Geotechnical and Geological Engineering, 39(2), 1171-1185.

    8. Lin, S. (2016). Linear and nonlinear approximation of spherical radial basis function networks. Journal of Complexity, 35, 86–101.

    9. Liang, L., Guo, W., Zhang, Y., Zhang, W., Li, L., & Xing, X. (2020). Radial basis function neural network for prediction of medium frequency sound absorption coefficient of composite structure open-cell aluminum foam. Applied Acoustics, 170, 107505.

    10. Mirjalili, S. (2019). Evolutionary radial basis function networks. Studies in Computational Intelligence, 1, 105–139.

    11. Silva, E.M., Maia, R.D., & Cabacinha, C.D. (2018). Bee-inspired RBF network for volume estimation of individual trees. Computers and Electronics in Agriculture, 152, 401–408.

    12. Shan, D., & Xu, X. (2017). Multi-label learning model based on multi-label radial basis function neural network and regularized extreme learning machine. Pattern Recognition, 30(9), 61-74.

    13. Tzortzis, G., & Likas, A. (2014). The MinMax k-Means clustering algorithm. Pattern Recognition, 47, 2505-2516.

    14. Wang, R., Li, Z., & Zhu, X. (2019). Deformation modulus prediction of hard rock mass using an improved RBF neural network model. Journal of Rock Mechanics and Geotechnical Engineering, 11(3), 531-540.

    15. Wang, Y., Liu, H., & Chen, L. (2021). An ensemble learning approach for predicting deformation modulus of hard rock mass based on improved RBF neural network. Journal of Computational Science, 54, 101078.

    16. Wu, J., Long, J., & Liu, M. (2015). Evolving RBF neural networks for rainfall prediction using hybrid particle swarm optimization and genetic algorithm. Neurocomputing, 148, 136-142.

    17. Yang, G., Wang, H., & Zhang, L. (2020). Improved prediction of deformation modulus of hard rock mass using hybrid RBF neural network optimized by bat algorithm. Journal of Petroleum Science and Engineering, 186, 106568.

    18. Yousef, W.A., & Kundu, S. (2014). Learning algorithms may perform worse with increasing training set size: Algorithm-data incompatibility. Computational Statistics and Data Analysis, 74, 181–197.

    19. Zhang, M., Wang, Y., & Zhang, X. (2018). Application of improved radial basis function neural network in predicting deformation modulus of hard rock mass. Applied Soft Computing, 68, 286-295.

    20. Zhang, G., Zhang, C., & Zhang, H. (2018). Improved K-means algorithm based on density canopy. Knowledge-Based Systems, 145, 289-297.

    21. Zhao, W.L., Deng, C.H., & Ngo, C.W. (2018). k-means: A revisit. Neurocomputing, 291, 195-206.