Bayesian optimization algorithm based support vector regression analysis for estimation of shear capacity of FRP reinforced concrete members

The Design of a Piecewise-Integrated Composite Bumper Beam with Machine-Learning Algorithms

In the present study, a piecewise-integrated composite bumper beam for passenger cars is proposed, and the design innovation process for a composite bumper beam regarding a bumper test protocol suggested by the Insurance Institute for Highway Safety is carried out with the help of machine learning models. Several elements in the bumper FE model have been assigned to be references in order to collect training data, which allow the machine learning model to study the method of predicting loading types for each finite element. Two-dimensional and three-dimensional implementations are provided by machine learning models, which determine the stacking sequences of each finite element in the piecewise-integrated composite bumper beam. It was found that the piecewise-integrated composite bumper beam, which is designed by a machine learning model, is more effective for reducing the possibility of structural failure as well as increasing bending strength compared to the conventional composite bumper beam. Moreover, the three-dimensional implementation produces better results compared with results from the two-dimensional implementation since it is preferable to choose loading-type information, which is achieved from surroundings when the target elements are located either at corners or junctions of planes, instead of using information that comes from the identical plane of target elements.

A novel model for rainfall prediction using hybrid stochastic-based Bayesian optimization algorithm

Rainfall forecasting is considered one of the key concerns in the meteorological department because it is related strongly to social as well as economic factors. But, because of modern context of climatic conditions and the intense activities of humans, the forecasting procedure of rainfall patterns becomes more problematic. Therefore, this paper proposes a novel timely and reliable rainfall prediction model using a hybrid stochastic Bayesian optimization approach (HS-BOA). The weather dataset containing different meteorological geographical features is provided as input to the introduced prediction method. Hybrid stochastic (HS) specifications are tuned by the Bayesian optimization algorithm (BOA) to upgrade the prediction accuracy. The weather data are initially preprocessed through the pipelines, namely, data separation, missing value prediction, weather condition cod separation, and normalization. After preprocessing, the highly correlated features are removed by correlation matrix using the Pearson correlation coefficient. Then, the most significant features which contribute more to predicting rainfall are selected through the feature selection process. At last, the suggested rainfall forecasting model accurately predicts rainfall using optimized parameters. The experimental analysis is performed, and for the proposed HS-BOA, MAE, RMSE, and COD, values attained for rainfall prediction are 0.513 mm, 59.90 mm, and 40.56 mm respectively. As a result, the proposed HS-BOA approach achieves minimum error rates with increased prediction accuracy than other existing approaches.

Application of Surface-Modified Nanoclay in a Hybrid Adsorption-Ultrafiltration Process for Enhanced Nitrite Ions Removal: Chemometric Approach vs. Machine Learning

Herein, we report the results of a study on combining adsorption and ultrafiltration in a single-stage process to remove nitrite ions from contaminated water. As adsorbent, a surface-modified nanoclay was employed (i.e., Nanomer^® I.28E, containing 25-30 wt. % trimethyl stearyl ammonium). Ultrafiltration experiments were conducted using porous polymeric membranes (Ultracel^® 10 kDa). The hybrid process of adsorption-ultrafiltration was modeled and optimized using three computational tools: (1) response surface methodology (RSM), (2) artificial neural network (ANN), and (3) support vector machine (SVM). The optimal conditions provided by machine learning (SVM) were found to be the best, revealing a rejection efficiency of 86.3% and an initial flux of permeate of 185 LMH for a moderate dose of the nanoclay (0.674% w/v). Likewise, a new and more retentive membrane (based on PVDF-HFP copolymer and halloysite (HS) inorganic nanotubes) was produced by the phase-inversion method, characterized by SEM, EDX, AFM, and FTIR techniques, and then tested under optimal conditions. This new composite membrane (PVDF-HFP/HS) with a thickness of 112 μm and a porosity of 75.32% unveiled an enhanced rejection efficiency (95.0%) and a lower initial flux of permeate (28 LMH). Moreover, molecular docking simulations disclosed the intermolecular interactions between nitrite ions and the functional moiety of the organonanoclay.

Machine learning models development for shear strength prediction of reinforced concrete beam: a comparative study

Fiber reinforced polymer (FPR) bars have been widely used as a substitutional material of steel reinforcement in reinforced concrete elements in corrosion areas. Shear resistance of FRP reinforced concrete element can be affected by concrete properties and transverse FRP stirrups. Hence, studying the shear strength (V_s) mechanism is one of the highly essential for pre-design procedure for reinforced concrete elements. This research examines the ability of three machine learning (ML) models called M5-Tree (M5), extreme learning machine (ELM), and random forest (RF) in predicting V_s of 112 shear tests of FRP reinforced concrete beam with transverse reinforcement. For generating the prediction matrix of the developed ML models, statistical correlation analysis was conducted to generate the suitable inputs models for V_s prediction. Statistical evaluation and graphical approaches were used to evaluate the efficiency of the proposed models. The results revealed that all the proposed models performed in general well for all the input combinations. However, ELM-M1 and M5-Tree-M5 models exhibited less accuracy performance in comparison with the other developed models. The study showed that the best prediction performance was revealed by M5 tree model using nine input parameters, with coefficient of determination (R²) and root mean square error (RMSE) equal to 0.9313 and 35.5083 KN, respectively. The comparison results also indicated that ELM and RF were performed significant results with a less slight performance than M5 model. The study outcome contributes to basic knowledge of investigating the impact of stirrups on V_s of FRP reinforced concrete beam with the potential of applying different computer aid models.

Establish a trend fuzzy information granule based short-term forecasting with long-association and k-medoids clustering

In the existing short-term forecasting methods of time series, two challenges are faced: capture the associations of data and avoid cumulative errors. For tackling these challenges, the fuzzy information granule based model catches our attention. The rule used in this model is fuzzy association rule (FAR), in which the FAR is constructed from a premise granule to a consequent granule at consecutive time periods, and then it describes the short-association in data. However, in real time series, another association, the association between a premise granule and a consequent granule at non-consecutive time periods, frequently exists, especially in periodical and seasonal time series. While the existing FAR can’t express such association. To describe it, the fuzzy long-association rule (FLAR) is proposed in this study. This kind of rule reflects the influence of an antecedent trend on a consequent trend, where these trends are described by fuzzy information granules at non-consecutive time periods. Thus, the FLAR can describe the long-association in data. Correspondingly, the existing FAR is called as fuzzy short-association rule (FSAR). Combining the existing FSAR with FLAR, a novel short-term forecasting model is presented. This model makes forecasting at granular level, and then it reduces the cumulative errors in short-term prediction. Note that the prediction results of this model are calculated from the available FARs selected by the k-medoids clustering based rule selection algorithm, therefore they are logical and accurate. The better forecasting performance of this model has been verified by comparing it with existing models in experiments.

Investigating the Bond Strength of FRP Rebars in Concrete under High Temperature Using Gene-Expression Programming Model

In recent times, the use of fibre-reinforced plastic (FRP) has increased in reinforcing concrete structures. The bond strength of FRP rebars is one of the most significant parameters for characterising the overall efficacy of the concrete structures reinforced with FRP. However, in cases of elevated temperature, the bond of FRP-reinforced concrete can deteriorate depending on a number of factors, including the type of FRP bars used, its diameter, surface form, anchorage length, concrete strength, and cover thickness. Hence, accurate quantification of FRP rebars in concrete is of paramount importance, especially at high temperatures. In this study, an artificial intelligence (AI)-based genetic-expression programming (GEP) method was used to predict the bond strength of FRP rebars in concrete at high temperatures. In order to predict the bond strength, we used failure mode temperature, fibre type, bar surface, bar diameter, anchorage length, compressive strength, and cover-to-diameter ratio as input parameters. The experimental dataset of 146 tests at various elevated temperatures were established for training and validating the model. A total of 70% of the data was used for training the model and remaining 30% was used for validation. Various statistical indices such as correlation coefficient (R), the mean absolute error (MAE), and the root-mean-square error (RMSE) were used to assess the predictive veracity of the GEP model. After the trials, the optimum hyperparameters were 150, 8, and 4 as number of chromosomes, head size and number of genes, respectively. Different genetic factors, such as the number of chromosomes, the size of the head, and the number of genes, were evaluated in eleven separate trials. The results as obtained from the rigorous statistical analysis and parametric study show that the developed GEP model is robust and can predict the bond strength of FRP rebars in concrete under high temperature with reasonable accuracy (i.e., R, RMSE and MAE 0.941, 2.087, and 1.620, and 0.935, 2.370, and 2.046, respectively, for training and validation). More importantly, based on the FRP properties, the model has been translated into traceable mathematical formulation for easy calculations.

Influence of Woven-Fabric Type on the Efficiency of Fabric-Reinforced Polymer Composites

The greatest advantage of fiber-reinforced composite materials is the freedom to tailor their strength and stiffness properties, while the most significant disadvantage consists in their high costs. Therefore, the design process and especially the optimization phase becomes an important step. The geometry of the fabric of each lamina as well as their stacking sequence need to be carefully defined, starting from some basic geometric variables. The input parameters are the widths and the heights of the tows, the laminate-stacking sequence and the gaps between two successive tows or the height of the neat matrix. This paper is a follow-up to a previous work on using and improving an in-house software called SOMGA (Satin Optimization with a Modified Genetic Algorithm), aimed to optimize the geometrical parameters of satin-reinforced multi-layer composites. The final goal is to find out the way in which various types of woven fabrics can affect the best possible solution to the problem of designing a composite material, able to withstand a given set of in-plane loads. The efficiency of the composite structure is evaluated by its ultimate strains using a fitness function that analyses and compares the mechanical behavior of different fabric-reinforced composites. Therefore, the ultimate strains corresponding to each configuration are considered intermediate data, being analyzed comparatively until obtaining the optimal values. When the software is running, for each analysis step, a set of intermediate values is provided. However, the users do not have to store these values, because the final result of the optimization directly provides the composite configuration with maximum efficiency, whose structural response meets the initially imposed loading conditions. To illustrate how the SOMGA software works, six different satin-woven-fabric-reinforced composites, starting from plain weave (satin 2/1/1), then satin 3/1/1, satin 4/1/1, satin 5/1/1, satin 5/2/1 and finally satin 5/3/1, were evaluated in the SOMGA interface. The results were rated against each other in terms of the composite efficiency and the case characterized by minimal reinforcement undulation (thinnest laminate) were highlighted.

Mechanical properties of coconut shell-based concrete: experimental and optimisation modelling

Excessive accumulation of waste materials has presented a serious environmental problem on a global scale. This has prompted many researchers to utilise agricultural, industrial, and by-product waste materials as the replacement of aggregate in the concrete matrix. In this present study, the prediction and optimisation of coconut shell (CA) content as the replacement of fine aggregate were evaluated based on the mechanical properties of the concrete (M30). Based on the suggested design array from the response surface methodology (RSM) model, experimental tests were carried out to achieve the goal of this study. The collected data was used to develop mathematical predictive equations using both GEP and RSM models. Analysis of variance (ANOVA) was also taken into account to appraise and verify the performance of the proposed models. Based on the results, the optimum content of replacing CA was 50%. In particular, the compressive, tensile, and flexural strength obtained after 28 days of curing were 46.2, 3.74, and 8.06 MPa, respectively, from the RSM model and 46.18, 3.85, and 7.99 MPa, respectively, from the GEP model. The obtained values were superior to those of the control concrete sample (43.12, 3.51 and 7.14 MPa, respectively). Beyond the optimum content, a loss in strength was observed. It was also found that both the GEP and RSM models exhibited high prediction accuracy with strong correlations (R² = 0.97 and 0.95, respectively). In addition, minimum prediction error (RMSE < 0.945 (RSM), RMSE < 1.62 (GEP)) was achieved, indicating that both models were robust and reliable for further prediction. It was concluded that CA could serve as an excellent strategic material to address several serious environmental issues.

Rock mass classification prediction model using heuristic algorithms and support vector machines: a case study of Chambishi copper mine

The rock mass is one of the key parameters in engineering design. Accurate rock mass classification is also essential to ensure operational safety. Over the past decades, various models have been proposed to evaluate and predict rock mass. Among these models, artificial intelligence (AI) based models are becoming more popular due to their outstanding prediction results and generalization ability for multiinfluential factors. In order to develop an easy-to-use rock mass classification model, support vector machine (SVM) techniques are adopted as the basic prediction tools, and three types of optimization algorithms, i.e., particle swarm optimization (PSO), genetic algorithm (GA) and grey wolf optimization (GWO), are implemented to improve the prediction classification and optimize the hyper-parameters. A database was assembled, consisting of 80 sets of real engineering data, involving four influencing factors. The three combined models are compared in accuracy, precision, recall, F₁ value and computational time. The results reveal that among three models, the GWO-SVC-based model shows the best classification performance by training. The accuracy of training and testing sets of GWO-SVC are 90.6250% (58/64) and 93.7500% (15/16), respectively. For Grades I, II, III, IV and V, the precision value is 1, 0.93, 0.90, 0.92, 0.83, the recall value is 1, 1, 0.93, 0.73, 0.83, and the F₁ value is 1, 0.96, 0.92, 0.81, 0.83, respectively. Sensitivity analysis is performed to understand the influence of input parameters on rock mass classification. It shows that the sensitive factor in rock mass quality is the RQD. Finally, the GWO-SVC is employed to assess the quality of rocks from the southeastern ore body of the Chambishi copper mine. Overall, the current study demonstrates the potential of using artificial intelligence methods in rock mass assessment, rendering far better results than the previous reports.

Prediction of FRCM–Concrete Bond Strength with Machine Learning Approach

Fibre-reinforced cement mortar (FRCM) has been widely utilised for the repair and restoration of building structures. The bond strength between FRCM and concrete typically takes precedence over the mechanical parameters. However, the bond behaviour of the FRCM–concrete interface is complex. Due to several failure modes, the prediction of bond strength is difficult to forecast. In this paper, effective machine learning models were employed in order to accurately predict the FRCM–concrete bond strength. This article employed a database of 382 test results available in the literature on single-lap and double-lap shear experiments on FRCM–concrete interfacial bonding. The compressive strength of concrete, width of concrete block, FRCM elastic modulus, thickness of textile layer, textile width, textile bond length, and bond strength of FRCM–concrete interface have been taken into consideration with popular machine learning models. The paper estimates the predictive accuracy of different machine learning models for estimating the FRCM–concrete bond strength and found that the GPR model has the highest accuracy with an R-value of 0.9336 for interfacial bond strength prediction. This study can be utilising in the estimation of bond strength to minimise the experimentation cost in minimum time.

Experimental Study of the Fatigue Performance of the Bonding Surfaces and Load-Bearing Capacity of a Large-Scale Severely Damaged Hollow Slab Strengthened by CFRP

In recent years, carbon fiber reinforced polymer (CFRP) has been widely used in bridge repair, retrofitting, rehabilitation and strengthening to improve the bearing capacity. Although many studies have been conducted to explore the strengthening efficiencies of CFRP, the test specimens were small and the results were difficult to apply to full-scale bridges. Investigations into the strengthening effects of CFRP on real life structures rely on field load tests (without damaging the structures), making it difficult to understand actual improvements in load carrying capacity and strengthening effect. Moreover, there have been few experimental studies on the fatigue performances of CFRP-strengthened structures, especially on the large-scale structures with real wheel moving loads. In this study, the feasibility and efficiency of CFRP strengthening and repair was investigated on a large-scale, prestressed concrete hollow slab decommissioned from a real-life concrete bridge. The hollow slab was first put through a destructive test to test the ultimate load-bearing capacity. Then, CFRP strips were installed on the surface of the severely damaged slab to repair and strengthen it. Fatigue load test—including the moving load test and single point sinusoidal load—and load-bearing capacity tests were conducted on the CFRP-strengthened hollow slab after the destructive test to evaluate the strengthening performance. This study could help us to understand the actual load-bearing capacities of severe damaged concrete structures strengthened by CFRP, reduce waste, save resources and improve the utilization of our infrastructures.