A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Prediction of the unconfined compressive strength of stabilised soil by Adaptive Neuro Fuzzy Inference System (ANFIS) and Non-Linear Regression (NLR)
This paper describes the effect of stabilizer content, curing time and moisture content on the UCS based upon 150 samples of stabilized soil. The results indicate an optimum value of lime or cement content which corresponds to the maximum UCS. The optimum lime content is dependent to moisture content, while the optimum cement content is independent of it. The increase of the UCS values due to increase of the stabilizer content in cement-stabilized samples is noticeably higher than lime-stabilized ones. The UCS values have also a direct and inverse relationship respectively with curing time and moisture content. UCS prediction is possible by incorporation of fuzzy logic in Adaptive Neuro Fuzzy Inference System (ANFIS). Fuzzy rules obtained from experiments are also incorporated into ANFIS to improve its learning capability. On the other hand, Non-Linear Regression (NLR) is employed more comfortably than ANFIS without need to rules. The results of sensitivity analysis indicate that increase of cement and moisture content have the most positive and negative effect, respectively, on the UCS values. It is observed a more satisfactory prediction of UCS in favour of ANFIS with 94% and 84% correlation coefficients between predicted and target values in cement and lime cases, respectively.
Prediction of the unconfined compressive strength of stabilised soil by Adaptive Neuro Fuzzy Inference System (ANFIS) and Non-Linear Regression (NLR)
This paper describes the effect of stabilizer content, curing time and moisture content on the UCS based upon 150 samples of stabilized soil. The results indicate an optimum value of lime or cement content which corresponds to the maximum UCS. The optimum lime content is dependent to moisture content, while the optimum cement content is independent of it. The increase of the UCS values due to increase of the stabilizer content in cement-stabilized samples is noticeably higher than lime-stabilized ones. The UCS values have also a direct and inverse relationship respectively with curing time and moisture content. UCS prediction is possible by incorporation of fuzzy logic in Adaptive Neuro Fuzzy Inference System (ANFIS). Fuzzy rules obtained from experiments are also incorporated into ANFIS to improve its learning capability. On the other hand, Non-Linear Regression (NLR) is employed more comfortably than ANFIS without need to rules. The results of sensitivity analysis indicate that increase of cement and moisture content have the most positive and negative effect, respectively, on the UCS values. It is observed a more satisfactory prediction of UCS in favour of ANFIS with 94% and 84% correlation coefficients between predicted and target values in cement and lime cases, respectively.
Prediction of the unconfined compressive strength of stabilised soil by Adaptive Neuro Fuzzy Inference System (ANFIS) and Non-Linear Regression (NLR)
Saadat, Mohsen (author) / Bayat, Meysam (author)
Geomechanics and Geoengineering ; 17 ; 80-91
2022-01-02
12 pages
Article (Journal)
Electronic Resource
Unknown
| British Library Conference Proceedings | 2005
| Taylor & Francis Verlag | 2019
| British Library Online Contents | 2017