Affiliations:
School of Civil Engineering, Architecture and Environment, Hubei University of Technology, Wuhan, China
Correspondence:
[*]
Corresponding author. Yong Miao, School of Civil Engineering, Architecture and Environment, Hubei University of Technology, Wuhan 430068, China. E-mail: 17868810692@163.com.
Abstract: The most significant parameter in groundwater movement in stones is capillary water absorption. Specifying the capillary water absorption (CWP) of rocks needs hard and laborious experimental work, while prediction models can reduce the cost and required time. To this aim, different rock specimens were gathered from various rocks. For the prediction processes, the hybrid adaptive neuro-fuzzy inference system (ANFIS) models also were proposed to determine the optimal value of two constituent parameters of the ANFIS, which the particle swarm optimization (PSO) and whale optimization algorithm (WOA) algorithm applied to the ANFIS for this aim. Results present that ANFIS processes have passable accomplishment in forecasting the CWA with R2 larger than 0.832 and 0.917 for the training and testing data, respectively, a good connection among actual and anticipated values. Considering developed models, the ANFIS model optimized with WOA performs better than another model in training and testing datasets. In the training dataset, the value of R2 and RRSE is 0.917 and 29.29% for the WOA-ANFIS model, while the PSO-ANFIS model is 0.911 and 30.50%, respectively. Overall, it is clear that WOA-ANFIS can be recognized as the proposed model, which shows its capability to find the optimal value of two constituent parameters of the ANFIS.
Keywords: Capillary water absorption, building stones, prediction, adaptive neuro-fuzzy inference system, hybrid ANFIS
