Table 1 Comparison of the work done in the research with similar literature.
Structure and topology | Key results | Refs. |
|---|---|---|
Implementation of feed-forward neural network, ENN, support vector machine and multi-linear regression for predicting the compressive strength of high-performance concrete | The values of R2 for the evaluation models were in the range of 0.9950 to 0.9853. The suggested model demonstrated significant accuracy and efficacy in forecasting the compressive strength of high-performance concrete | |
Compressive strength of palm oil fuel ash concrete was predicted using six concrete parameters as inputs of the elaborated model. Four machine learning techniques were compared in many relationships | ANN and ANN with combined inputs outperformed PSO and GA in terms of predictive accuracy. Further, ANN with combined inputs demonstrated a relatively higher performance level than ANN | |
An ANN was provided to predict pervious concrete properties. Accurate prediction of permeability and compressive strength were attained | Predictive accuracies of R2 = 0.98 for permeability and R2 = 0.97 for compressive strength were achieved. The ANN models offered viable alternative to tedious lab tests for FRC assessment | |
A model was developed to predict the compressive strength and crushing strain of concrete confined with fiber-reinforced polymers utilizing neural networks and regression techniques | The model utilizing neural networks exhibited a high level of predictive accuracy, and the findings indicated that employing neural networks to evaluate the compressive strength and crushing strain of FRP-confined concrete is both effective and advantageous | |
Two distinct models of ANN and two separate Adaptive Network-based Fuzzy Inference Systems were developed to forecast the compressive strength of seven various cement mortar samples, which included combinations of pumice and/or diatomite, assessed over multiple days | Pumice and diatomite contribute to compressive strength of mortars in later ages. Cement mortars' compressive strength could be estimated with a very small error and short time with ANN and ANFIS models. ANN models showed relatively better prediction performance compared to ANFIS models | |
Non-linear ultrasonic and ANN were used for non-destructive evaluation of the damages in concrete. The time-domain signals of the received ultrasonic waves were used | Voltage of input pulse and peak-to-peak voltage were more important than the average pulse velocity in prediction of damage | |
Agro-waste was utilized as a substitute in cement to reduce landfill problems. The best-fit statistical model was identified for predicting compressive strength of concrete | De-oiled Earth and rice husk ash had good pozzolanic reaction to be used as a cementitious material. The R2 values derived from the regression and ANN methodologies were 0.6561 and 0.9673 for the 7-day period, while for the 28-day period, the values were 0.6441 and 0.9636, correspondingly | |
The modeling of air chamber pressure control in slurry shield tunneling was developed. A predictive control system based on ENN model for air pressure was proposed. PSO algorithm was implemented to improve the learning capability of ENN model | The implementation of a predictive control system utilizing a PSO-based ENN model demonstrated significant potential for improving face stability during slurry shield tunneling operations | |
Various AI methods had been employed to predict the compressive strength of concrete containing sugarcane bagasse ash. AI models trained on concrete datasets with sugarcane bagasse ash, fly ash, slag, and none to assess their impact | Sugarcane bagasse ash could be used as a partial replacement of cement in concrete mixtures. The multi-objective genetic algorithms could be employed to optimise sugarcane bagasse ash concrete for cost-effective compressive strength | |
A comprehensive dataset for modeling the compressive strength of Carbon Fiber-Reinforced Polymer Confined-Concrete- Confined Circular Concrete was gathered. The predicting the compressive strength of CFRP- CC specimens was based on using data-driven methods | The model demonstrated exceptional predictive performance, as indicated by an outstanding R2 of 0.9847, which exceeds that of alternative methodologies | |
Evaluating the performance of Developed African Vulture Optimization Algorithm-Elman Neural Networks by considering different input parameters in predicting the self-compacting concrete compressive strength | In predicting both 7-day and 28-day compressive strength, networks built with 140 parameters have a 74.54 and 70.44% improvement in test error over 8-parameter networks, respectively, which directly affects this effect | Current paper |
