Assessment of different machine learning techniques in predicting the compressive strength of self-compacting concrete

The compressive strength of self-compacting concrete (SCC) needs to be determined during the construction design process. This paper shows that the compressive strength of SCC (CS of SCC) can be successfully predicted from mix design and curing age by a machine learning (ML) technique named the Extreme Gradient Boosting (XGB) algorithm, including non-hybrid and hybrid models. Nine ML techniques, such as Linear regression (LR), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Decision Trees (DTR), Random Forest (RF), Gradient Boosting (GB), and Artificial Neural Network using two training algorithms LBFGS and SGD (denoted as ANN_LBFGS and ANN_SGD), are also compared with the XGB model. Moreover, the hybrid models of eight ML techniques and Particle Swarm Optimization (PSO) are constructed to highlight the reliability and accuracy of SCC compressive strength prediction by the XGB_PSO hybrid model. The highest number of SCC samples available in the literature is collected for building the ML techniques. Compared with previously published works’ performance, the proposed XGB method, both hybrid and non-hybrid models, is the most reliable and robust of the examined techniques, and is more accurate than existing ML methods (R² = 0.9644, RMSE = 4.7801, and MAE = 3.4832). Therefore, the XGB model can be used as a practical tool for engineers in predicting the CS of SCC.     

Optimization of machine learning models for predicting the compressive strength of fiber-reinforced self-compacting concrete

Fiber-reinforced self-compacting concrete (FRSCC) is a typical construction material, and its compressive strength (CS) is a critical mechanical property that must be adequately determined. In the machine learning (ML) approach to estimating the CS of FRSCC, the current research gaps include the limitations of samples in databases, the applicability constraints of models owing to limited mixture components, and the possibility of applying recently proposed models. This study developed different ML models for predicting the CS of FRSCC to address these limitations. Artificial neural network, random forest, and categorical gradient boosting (CatBoost) models were optimized to derive the best predictive model with the aid of a 10-fold cross-validation technique. A database of 381 samples was created, representing the most significant FRSCC dataset compared with previous studies, and it was used for model development. The findings indicated that CatBoost outperformed the other two models with excellent predictive abilities (root mean square error of 2.639 MPa, mean absolute error of 1.669 MPa, and coefficient of determination of 0.986 for the test dataset). Finally, a sensitivity analysis using a partial dependence plot was conducted to obtain a thorough understanding of the effect of each input variable on the predicted CS of FRSCC. The results showed that the cement content, testing age, and superplasticizer content are the most critical factors affecting the CS.     

A novel ensemble model for predicting the performance of a novel vertical slot fishway

We investigate the performance of a novel vertical slot fishway by employing finite volume and surrogate models. Multiple linear regression, multiple log equation regression, gene expression programming, and combinations of these models are employed to predict the maximum turbulence, maximum velocity, resting area, and water depth of the middle pool in the fishway. The statistical parameters and error terms, including the coefficient of determination, root mean square error, normalized square error, maximum positive and negative errors, and mean absolute percentage error were employed to evaluate and compare the accuracy of the models. We also conducted a parametric study. The independent variables include the opening between baffles (OBB), the ratio of the length of the large and small baffles, the volume flow rate, and the angle of the large baffle. The results show that the key parameters of the maximum turbulence and velocity are the volume flow rate and OBB.     

Predicting the settlement of geosynthetic-reinforced soil foundations using evolutionary artificial intelligence technique

In order to ensure safe and sustainable design of geosynthetic-reinforced soil foundation (GRSF), settlement prediction is a challenging task for practising civil/geotechnical engineers. In this paper, a new hybrid technique for predicting the settlement of GRSF has been proposed based on the combination of evolutionary algorithm, that is, grey-wolf optimisation (GWO) and artificial neural network (ANN), abbreviated as ANN-GWO model. For this purpose, the reliable pertinent data were generated through numerical simulations conducted on validated large-scale 3-D finite element model. The predictive power of the model was assessed using various well-established statistical indices, and also validated against several independent scientific studies as reported in literature. Furthermore, the sensitivity analysis was conducted to examine the robustness and reliability of the model. The results as obtained have indicated that the developed hybrid ANN-GWO model can estimate the maximum settlement of GRSF under service loads in a reliable and intelligent way, and thus, can be deployed as a predictive tool for the preliminary design of GRSF. Finally, the model was translated into functional relationship which can be executed without the need of any expensive computer-based program.     

基于SPSS回归分析的锂渣混凝土抗压强度预测模型

利用SPSS软件的逐步回归分析法、多元非线性回归法建立锂渣混凝土的强度预测模型,分析各模型的残差图、预测值与试验值的对比,并结合均方根误差、平均绝对误差、平均绝对百分比误差和模型可决系数值对各模型的精确度等进行综合评价,最终确定出较优的锂渣混凝土强度预测模型。结果表明:水胶比、锂渣掺量和减水剂掺量对锂渣混凝土强度的影响十分显著;经残差分析和95%预测值区间检验,5个建议模型都有较好的精确度;经综合评价建议最佳的锂渣混凝土强度预测模型是以水泥强度、胶水比、锂渣掺量和减水剂掺量为自变量的非线性回归方程,其相应的可决系数R²=0.920,均方根误差为3.684,平均绝对误差为3.15,平均绝对百分比误差为5.44。     

Assessment of evolutionary algorithms in predicting non-deposition sediment transport

In this article, the densimetric Froude number of the flow is estimated using the parameters of volumetric sediment concentration (C_V), the relative depth of flow (d/R), dimensionless particle number (D_gr) and the overall sediment friction factor (λ_s). The particle swarm optimization (PSO) and imperialist competitive algorithms (ICA) were used to estimate the densimetric Froude number. To study the effects of sediment transport parameters on the densimetric Froude number, six different models are presented. The PSO algorithm with root mean square error (RMSE) = 0.014 and mean absolute percentage error (MAPE) = 5.1% present the results with a relatively good accuracy. The accuracy of the results presented for the selected model by the ICA algorithm is also in the form of RMSE = 0.007 and MAPE = 5.6%. Although both algorithms return good results in estimating the densimetric Froude number for the selected model, it should be mentioned that for all the six presented models ICA returns better results than PSO.     

基于长短时记忆网络的盾构机刀盘扭矩实时预测

盾构机是土质隧道开挖的优质工程机械,被广泛应用在地铁建设。刀盘扭矩是保证盾构正常推进的关键参数,能被精确实时预测对预防灾难事故、确保施工正常推进具有极高的指导意义。针对现有扭矩预测多为计算平均值的问题,提出一种基于长短时记忆(Long-Short Term Memory,LSTM)网络的扭矩实时预测模型。首先通过分析盾构机状态参数与扭矩的相关性,选择一组关键状态参数,降低输入维度;然后建立LSTM扭矩预测模型;最后利用该模型在归一化后的实际数据集上进行验证并与BP网络模型对比。试验结果表明,该模型在测试集上均方差为0.002 81,平均绝对误差为0.036 7,均优于BP网络;该模型具有良好的预测能力与泛化性能,能够很好地拟合关键状态参数与刀盘扭矩之间的非线性关系。     

Prediction of shield tunneling-induced ground settlement using machine learning techniques

Predicting the tunneling-induced maximum ground surface settlement is a complex problem since the settlement depends on plenty of intrinsic and extrinsic factors. This study investigates the efficiency and feasibility of six machine learning (ML) algorithms, namely, back-propagation neural network, wavelet neural network, general regression neural network (GRNN), extreme learning machine, support vector machine and random forest (RF), to predict tunneling-induced settlement. Field data sets including geological conditions, shield operational parameters, and tunnel geometry collected from four sections of tunnel with a total of 3.93 km are used to build models. Three indicators, mean absolute error, root mean absolute error, and coefficient of determination the (R²) are used to demonstrate the performance of each computational model. The results indicated that ML algorithms have great potential to predict tunneling-induced settlement, compared with the traditional multivariate linear regression method. GRNN and RF algorithms show the best performance among six ML algorithms, which accurately recognize the evolution of tunneling-induced settlement. The correlation between the input variables and settlement is also investigated by Pearson correlation coefficient.     

Predicting lateral displacement caused by seismic liquefaction and performing parametric sensitivity analysis: Considering cumulative absolute velocity and fine content

Lateral displacement due to liquefaction (D_H) is the most destructive effect of earthquakes in saturated loose or semi-loose sandy soil. Among all earthquake parameters, the standardized cumulative absolute velocity (CAV₅) exhibits the largest correlation with increasing pore water pressure and liquefaction. Furthermore, the complex effect of fine content (FC) at different values has been studied and demonstrated. Nevertheless, these two contexts have not been entered into empirical and semi-empirical models to predict D_H. This study bridges this gap by adding CAV₅ to the data set and developing two artificial neural network (ANN) models. The first model is based on the entire range of the parameters, whereas the second model is based on the samples with FC values that are less than the 28% critical value. The results demonstrate the higher accuracy of the second model that is developed even with less data. Additionally, according to the uncertainties in the geotechnical and earthquake parameters, sensitivity analysis was performed via Monte Carlo simulation (MCS) using the second developed ANN model that exhibited higher accuracy. The results demonstrated the significant influence of the uncertainties of earthquake parameters on predicting D_H.     

Development of room temperature and relative humidity linear parametric models for an open office using BMS data

This study investigates Box-Jenkins (BJ), autoregressive with external inputs (ARX), autoregressive moving average with external inputs (ARMAX) and output error (OE) models to identify the thermal behaviour of an office positioned in a modern commercial building in London. These models can all be potentially used for improving the performance of the thermal environment control system. External and internal climate data, recorded over the summer, autumn and winter seasons, were used to build and validate the models. The paper demonstrates the potential of using linear parametric models to predict room temperature and relative humidity for different time scales (30 min or 2 h ahead). The prediction performance is evaluated using the criteria of goodness of fit, coefficient of determination, mean absolute error and mean squared error between predicted model output and real measurements. The results demonstrate that all models provide reasonably good predictions but the BJ model outperforms the ARMAX and ARX models.     

Prediction of mode I fracture toughness of rock using linear multiple regression and gene expression programming

Prediction of mode I fracture toughness(KIC) of rock is of significant importance in rock engineering analyses. In this study, linear multiple regression(LMR) and gene expression programming(GEP)methods were used to provide a reliable relationship to determine mode I fracture toughness of rock. The presented model was developed based on 60 datasets taken from the previous literature. To predict fracture parameters, three mechanical parameters of rock mass including uniaxial compressive strength(...     

Realtime prediction of hard rock TBM advance rate using temporal convolutional network (TCN) with tunnel construction big data

Real-time dynamic adjustment of the tunnel bore machine (TBM) advance rate according to the rock-machine interaction parameters is of great significance to the adaptability of TBM and its efficiency in construction. This paper proposes a real-time predictive model of TBM advance rate using the temporal convolutional network (TCN), based on TBM construction big data. The prediction model was built using an experimental database, containing 235 data sets, established from the construction data from the Jilin Water-Diversion Tunnel Project in China. The TBM operating parameters, including total thrust, cutterhead rotation, cutterhead torque and penetration rate, are selected as the input parameters of the model. The TCN model is found outperforming the recurrent neural network (RNN) and long short-term memory (LSTM) model in predicting the TBM advance rate with much smaller values of mean absolute percentage error than the latter two. The penetration rate and cutterhead torque of the current moment have significant influence on the TBM advance rate of the next moment. On the contrary, the influence of the cutterhead rotation and total thrust is moderate. The work provides a new concept of real-time prediction of the TBM performance for highly efficient tunnel construction.     

基于MPS方法的自密实混凝土流动数值模拟

采用三维移动粒子半隐式(MPS)法来分析和预测自密实混凝土(SCC)的流动性,对L形箱试验进行数值模拟,对比分析了数值模拟与试验中SCC的流动形态和流动时间.在此基础上,研究了L形箱试验和V形漏斗试验中塑性黏度对SCC流动性的影响,并建立了SCC密度和塑性黏度与V形漏斗试验下落时间的关系.结果表明:使用基于Bingham流变模型的MPS方法可以较好地模拟SCC的流动特性,且精度较高;MPS方法可以对SCC在施工过程中的流动时间和流动形态进行模拟预测,结合流变参数分析可以为SCC工程的设计和施工实践提供参考依据.     

Hybrid intelligent inference model for enhancing prediction accuracy of scour depth around bridge piers

Bridge-pier scouring is a main cause of bridge failures. Thus, accurately predicting the scour depth around bridge piers is critical, both to specify adequate depths for new bridge foundations and to assess/monitor the safety of existing bridges. This study proposes a novel artificial intelligence (AI) model, the intelligent fuzzy radial basis function neural network inference model (IFRIM), to estimate future scour depth around bridge piers. IFRIM is a hybrid of the radial basis function neural network (RBFNN), fuzzy logic (FL), and the artificial bee Cclony (ABC) algorithm. In the IFRIM, FL is used to handle the uncertainties in input information, RBFNN is used to handle the fuzzy input–output mapping relationships, and the ABC search engine employs optimisation to identify the most suitable tuning parameters for RBFNN and FL based on minimal error estimation. A 10-fold cross-validation method finds that the IFRIM model achieves at least 21% and 14.5% reductions in root mean square error and mean absolute error values, respectively, compared with other AI techniques. Study results support the IFRIM as a promising new tool for civil engineers to predict future scour depth around bridge piers.     

型钢自密实混凝土叠合剪力墙恢复力模型

为了研究型钢自密实混凝土叠合剪力墙的抗震性能以及恢复力特性,对1个全现浇型钢普通混凝土剪力墙试件和5个型钢自密实混凝土叠合剪力墙试件进行了低周反复加载试验。通过分析试验所得的剪力墙的水平荷载-位移骨架曲线,建议将其简化为以屈服点、峰值荷载点、极限点为特征点的三折线模型,并根据试验现象以及理论分析结果给出骨架曲线中各特征点参数的计算方法。对试验所得滞回曲线进行回归分析,得到了墙体的抗侧刚度退化规律,据此得出适用于型钢自密实混凝土叠合剪力墙的恢复力模型,并用试验得到的骨架曲线与滞回曲线进行验证。结果表明,由此恢复力模型求得的计算曲线与试验曲线吻合程度较好,该模型可以较好地反映型钢自密实混凝土叠合剪力墙的恢复力特性,可用于计算地震作用下以受弯破坏为主的型钢自密实混凝土叠合剪力墙的荷载-位移滞回曲线。     

小波神经网络在水库流量预测中的应用

针对河南省某水库监测点实测的1991~2013年每月的平均流量样本进行归一化处理作为训练样本,构建了使用Morlet、Mexican hat以及高斯一阶导数小波基函数小波神经网络的预测模型实现对2014年的月平均流量的预测,并通过均方误差(MSE)和平均绝对误差(MAE)两项指标对每种网络的预测结果进行评价,从而选择较好的小波基函数作为小波神经网络的隐含层传递函数。研究表明,采用Morlet小波作为神经网络的隐含层基函数对该水库的月平均流量的预测效果要好于其他两种神经网络。     

Estimating dynamic origin–destination demand: A hybrid framework using license plate recognition data

This article proposes a hybrid framework for estimating dynamic origin–destination (OD) demand that fully exploits the information available in license plate recognition (LPR) data. A Bayesian path reconstruction model is initially developed to replenish the lost information resulting from the recognition error and insufficient coverage rate of the LPR system. The link flows, initial OD demand, left‐turning flows, and partial path flows are derived based on the reconstructed data. Subsequently, with the information derived, a two‐step ordinary least squares (OLS) OD estimation model is formulated, which incorporates the output from the Bayesian model and coestimates the OD demand and assignment matrix. The proposed framework is qualitatively validated using the real‐world LPR data collected from Langfang City, Hebei Province, China, and is quantitatively validated using the synthesized simulation data for the simplified road network of Langfang. The results show that the proposed model can estimate OD demand distribution with a mean absolute percentage error (MAPE) of about 30%. We also tested the model with different LPR coverage rates, with results showing that an LPR coverage rate of over 50% is required to obtain reasonable results.     

Estimation of convergence of a high-speed railway tunnel in weak rocks using an adaptive neuro-fuzzy inference system(ANFIS) approach

Estimation of tunnel diameter convergence is a very important issue for tunneling construction,especially when the new Austrian tunneling method(NATM) is adopted.For this purpose,a systematic convergence measurement is usually implemented to adjust the design during the whole construction,and consequently deadly hazards can be prevented.In this study,a new fuzzy model capable of predicting the diameter convergences of a high-speed railway tunnel was developed on the basis of adaptive neuro-fuzzy inference system(ANFIS) approach.The proposed model used more than 1 000 datasets collected from two different tunnels,i.e.Daguan tunnel No.2 and Yaojia tunnel No.1,which are part of a tunnel located in Hunan Province,China.Six Takagi-Sugeno fuzzy inference systems were constructed by using subtractive clustering method.The data obtained from Daguan tunnel No.2 were used for model training,while the data from Yaojia tunnel No.1 were employed to evaluate the performance of the model.The input parameters include surrounding rock masses(SRM) rating index,ground engineering conditions(GEC) rating index,tunnel overburden(H),rock density(?),distance between monitoring station and working face(D),and elapsed time(T).The model’s performance was assessed by the variance account for(VAF),root mean square error(RMSE),mean absolute percentage error(MAPE) as well as the coefficient of determination(R2) between measured and predicted data as recommended by many researchers.The results showed excellent prediction accuracy and it was suggested that the proposed model can be used to estimate the tunnel convergence and convergence velocity.     

Developing cation exchange capacity and soil index properties relationships using a neuro-fuzzy approach

Artificial intelligence methods are employed to predict cation exchange capacity (CEC) from five different soil index properties, namely specific surface area (SSA), liquid limit, plasticity index, activity (ACT), and clay fraction (CF). Artificial neural networks (ANNs) analyses were first employed to determine the most related index parameters with cation exchange capacity. For this purpose, 40 datasets were employed to train the network and 10 datasets were used to test it. The ANN analyses were conducted with 15 different input vector combinations using same datasets. As a result of this investigation, the ANN analyses revealed that SSA and ACT are the most effective parameters on the CEC. Next, based upon these most effective input parameters, the fuzzy logic (FL) model was developed for the CEC. In the developed FL model, triangular membership functions were employed for both the input (SSA and ACT) variables and the output variable (CEC). A total of nine Mamdani fuzzy rules were deduced from the datasets, used for the training of the ANN model. Minimization (min) inferencing, maximum (max) composition, and centroid defuzzification methods are employed for the constructed FL model. The developed FL model was then tested against the remaining datasets, which were also used for testing the ANN model. The prediction results are satisfactory with a determination coefficient, R ² = 0.94 and mean absolute error, (MAE) = 7.1.     

Prediction of blast-induced ground vibration using artificial neural network

An attempt has been made to evaluate and predict the blast-induced ground vibration and frequency by incorporating rock properties, blast design and explosive parameters using the artificial neural network (ANN) technique. A three-layer, feed-forward back-propagation neural network having 15 hidden neurons, 10 input parameters and two output parameters were trained using 154 experimental and monitored blast records from one of the major producing surface coal mines in India. Twenty new blast data sets were used for the validation and comparison of the peak particle velocity (PPV) and frequency by ANN and other predictors. To develop more confidence in the proposed method, same data sets have also been used for the prediction of PPV by commonly used vibration predictors as well as by multivariate regression analysis (MVRA). Results were compared based on correlation and mean absolute error (MAE) between monitored and predicted values of PPV and frequency.