An ensemble method for predicting the mechanical properties of strain hardening cementitious composites

Abstract In recent years, Artificial Neural Networks (ANN) models have proven effective in learning to predict material properties only from data. In the study of strain-hardening cementitious composites (SHCC), conducting laboratory experimentation to collect this data is expensive, and to reach reliable accuracy, tens of thousands of instances may be required. In this paper, a forest deep neural network (FDNN) ensemble model is presented. The proposed model combines random forest regressors with artificial neural networks to predict the cementitious composite’s mechanical properties. The FDNN model can provide reliable predictions after training on relatively small datasets of different classes of SHCC. The FDNN outperforms previously developed artificial neural networks model architectures build specifically for polyvinyl alcohol (PVA) based SHCC. The model achieves an average root means square error of 0.08, and its predictions for several specimens are provided and discussed in this work.

[1]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[2]  C. Leung,et al.  Strength Improvement of Strain-Hardening Cementitious Composites with Ultrahigh-Volume Fly Ash , 2017 .

[3]  Erwan Scornet,et al.  Random Forests and Kernel Methods , 2015, IEEE Transactions on Information Theory.

[4]  Zhongqing Su,et al.  Application of Kernel Density Estimation in Lamb Wave-Based Damage Detection , 2012 .

[5]  Shane Legg,et al.  Human-level control through deep reinforcement learning , 2015, Nature.

[6]  Joachim Denzler,et al.  Deep Learning on Small Datasets without Pre-Training using Cosine Loss , 2019, 2020 IEEE Winter Conference on Applications of Computer Vision (WACV).

[7]  C. Leung,et al.  Development of Multiscale Fiber-Reinforced Engineered Cementitious Composites with PVA Fiber and CaCO3 Whisker , 2018, Journal of Materials in Civil Engineering.

[8]  Zhishen Wu,et al.  Experimental study on effect of resin matrix in basalt fiber reinforced polymer composites under static and fatigue loading , 2020 .

[9]  Synho Do,et al.  How much data is needed to train a medical image deep learning system to achieve necessary high accuracy , 2015, 1511.06348.

[10]  Christopher K.Y. Leung,et al.  Effect of graphene oxide on the mechanical behavior of strain hardening cementitious composites , 2016 .

[11]  Jerome P. Lynch,et al.  Development of thermally adaptive Engineered Cementitious Composite for passive heat storage , 2014 .

[12]  Shijie Cheng,et al.  Artificial Intelligence to Power the Future of Materials Science and Engineering , 2020, Adv. Intell. Syst..

[13]  Radu Horaud,et al.  A Comprehensive Analysis of Deep Regression , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[14]  Yoshua Bengio,et al.  Practical Recommendations for Gradient-Based Training of Deep Architectures , 2012, Neural Networks: Tricks of the Trade.

[15]  V. Li,et al.  Mechanical performance of ECC with high-volume fly ash after sub-elevated temperatures , 2015 .

[16]  Mohamed Maalej,et al.  Tensile strain hardening behaviour of hybrid steel-polyethylene fibre reinforced cementitious composites , 2009 .

[17]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[18]  De‐Zhu Li,et al.  Genetic structure and differentiation in Dendrocalamus sinicus (Poaceae: Bambusoideae) populations provide insight into evolutionary history and speciation of woody bamboos , 2018, Scientific Reports.

[19]  Victor C. Li,et al.  High-Early-Strength Engineered Cementitious Composites for Fast, Durable Concrete Repair—Material Properties , 2011 .

[20]  Mohamed Maalej,et al.  Tensile Behavior of Cement‐Based Composites with Random Discontinuous Steel Fibers , 1996 .

[21]  Donald C. Wunsch,et al.  Classical Neural Networks , 2010 .

[22]  V. Li,et al.  Effect of morphological parameters of natural sand on mechanical properties of engineered cementitious composites , 2019, Cement and Concrete Composites.

[23]  Mohamed Lachemi,et al.  Frost resistance and microstructure of Engineered Cementitious Composites: Influence of fly ash and micro poly-vinyl-alcohol fiber , 2012 .

[24]  James H. Garrett,et al.  Knowledge-Based Modeling of Material Behavior with Neural Networks , 1992 .

[25]  V. Li,et al.  Influence of aggregate type and size on ductility and mechanical properties of engineered cementitious composites , 2009 .

[26]  Zhishen Wu,et al.  Temperature effect on fatigue behavior of basalt fiber‐reinforced polymer composites , 2018, Polymer Composites.

[27]  H. K. D. H. Bhadeshia,et al.  Neural Networks in Materials Science , 1999 .

[28]  Matthias Hein,et al.  Why ReLU Networks Yield High-Confidence Predictions Far Away From the Training Data and How to Mitigate the Problem , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[29]  Victor C. Li,et al.  Engineered Cementitious Composites (ECC) Material, Structural, and Durability Performance , 2008 .

[30]  Zongjin Li,et al.  Performance of novel thermal energy storage engineered cementitious composites incorporating a paraffin/diatomite composite phase change material , 2014 .

[31]  Lin Ye,et al.  Artificial neural network based mechanical and electrical property prediction of engineered cementitious composites , 2018, Construction and Building Materials.

[32]  俊一 甘利.  5分で分かる!? 有名論文ナナメ読み：Jacot, Arthor, Gabriel, Franck and Hongler, Clement : Neural Tangent Kernel : Convergence and Generalization in Neural Networks , 2020 .

[33]  Tuan Nguyen,et al.  Deep neural network with high‐order neuron for the prediction of foamed concrete strength , 2018, Comput. Aided Civ. Infrastructure Eng..

[34]  C. Leung,et al.  Use of high strength Strain-Hardening Cementitious Composites for flexural repair of concrete structures with significant steel corrosion , 2018 .

[35]  Alberto Caballero,et al.  Artificial Intelligence Systems in Micromechanics , 2004, AIAI.

[36]  Dabal Pedamonti,et al.  Comparison of non-linear activation functions for deep neural networks on MNIST classification task , 2018, ArXiv.

[37]  Tianwei Yu,et al.  A Deep Neural Network Model using Random Forest to Extract Feature Representation for Gene Expression Data Classification , 2018, Scientific Reports.

[38]  Hashim Abdul Razak,et al.  Flexural behavior of engineered cementitious composite (ECC) slabs with polyvinyl alcohol fibers , 2015 .

[39]  C. Leung,et al.  Graphene oxide modified Strain Hardening Cementitious Composites with enhanced mechanical and thermal properties by incorporating ultra-fine phase change materials , 2019, Cement and Concrete Composites.

[41]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[42]  Seyedali Mirjalili,et al.  Novel Feature Selection and Voting Classifier Algorithms for COVID-19 Classification in CT Images , 2020, IEEE Access.

[43]  Jaehoon Lee,et al.  Wide neural networks of any depth evolve as linear models under gradient descent , 2019, NeurIPS.

[44]  H. Mihashi,et al.  A review on durability properties of strain hardening fibre reinforced cementitious composites (SHFRCC) , 2007 .

[45]  V. Li,et al.  Effect of Fly Ash and PVA Fiber on Microstructural Damage and Residual Properties of Engineered Cementitious Composites Exposed to High Temperatures , 2011 .

[46]  V. Li,et al.  Scale-linking model of self-healing and stiffness recovery in Engineered Cementitious Composites (ECC) , 2019, Cement and Concrete Composites.

[47]  Wu Gang,et al.  ADVANCEMENT OF STRUCTURAL SAFETY AND SUSTAINABILITY WITH BASALT FIBER REINFORCED POLYMERS , 2012 .

[48]  V. Li,et al.  Thermal-mechanical behaviors of CFRP-ECC hybrid under elevated temperatures , 2017 .

[49]  S. Pourfalah,et al.  Behaviour of engineered cementitious composites and hybrid engineered cementitious composites at high temperatures , 2018 .

[50]  Elad Hoffer,et al.  Train longer, generalize better: closing the generalization gap in large batch training of neural networks , 2017, NIPS.

[51]  Sepp Hochreiter,et al.  Self-Normalizing Neural Networks , 2017, NIPS.

[52]  Tadashi Saito,et al.  Interface tailoring for strain-hardening polyvinyl alcohol-engineered cementitious composite (PVA-ECC) , 2002 .

[53]  Leo Breiman,et al.  Stacked regressions , 2004, Machine Learning.

[54]  C. Leung,et al.  Methods of modifying the brittle behavior of cementitious composites , 2010 .

[55]  M. Bessa,et al.  Bayesian Machine Learning in Metamaterial Design: Fragile Becomes Supercompressible , 2019, Advanced materials.

[56]  Burak Felekoğlu,et al.  Influence of matrix flowability, fiber mixing procedure, and curing conditions on the mechanical performance of HTPP-ECC , 2014 .

[57]  C. Hung,et al.  Innovative ECC jacketing for retrofitting shear-deficient RC members , 2016 .

[58]  P. Pan,et al.  Effect of water to binder ratio and sand to binder ratio on shrinkage and mechanical properties of High-strength Engineered Cementitious Composite , 2019, Construction and Building Materials.

[59]  Hwai Chung Wu,et al.  Matrix design for pseudo-strain-hardening fibre reinforced cementitious composites , 1995 .

[60]  Swee Leong Kok,et al.  Review of potential structural applications of hybrid fiber Engineered Cementitious Composites , 2012 .

[61]  Chongshi Gu,et al.  Influence of micro-cracking on the permeability of engineered cementitious composites , 2016 .

[62]  Jinping Ou,et al.  An economical ultra-high ductile engineered cementitious composite with large amount of coarse river sand , 2019, Construction and Building Materials.

[63]  C. Leung,et al.  Tensile performance and impact resistance of Strain Hardening Cementitious Composites (SHCC) with recycled fibers , 2018 .

[64]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[65]  C. Leung,et al.  Development of pseudo-ductile permanent formwork for durable concrete structures , 2010 .

[66]  Hui Li,et al.  The State of the Art of Data Science and Engineering in Structural Health Monitoring , 2019, Engineering.

[67]  Trevor Hastie,et al.  An Introduction to Statistical Learning , 2013, Springer Texts in Statistics.

[68]  R. Ranade,et al.  Influence of high temperatures on the residual mechanical properties of a hybrid fiber-reinforced strain-hardening cementitious composite , 2019, Construction and Building Materials.

[69]  Victor C. Li,et al.  Engineered Cementitious Composites with High-Volume Fly Ash , 2007 .

[70]  Shunzhi Qian,et al.  Self-healing behavior of strain hardening cementitious composites incorporating local waste materials , 2009 .

[71]  Björn Johannesson,et al.  A review : Self-healing in cementitious materials and engineered cementitious composite as a self-healing material , 2012 .

[72]  Ahmet Çavdar,et al.  A study on the effects of high temperature on mechanical properties of fiber reinforced cementitious composites , 2012 .

[73]  J H Garrett,et al.  KNOWLEDGE-BASED MODELLING OF MATERIAL BEHAVIOUR WITH NEURAL NETWORKS , 1991 .

[74]  Gaurav Vats,et al.  Mutual Insight on Ferroelectrics and Hybrid Halide Perovskites: A Platform for Future Multifunctional Energy Conversion , 2019, Advanced materials.

[75]  Babita Saini,et al.  Performance and composition analysis of engineered cementitious composite (ECC) – A review , 2019, Journal of Building Engineering.

[76]  Khandaker M. A. Hossain,et al.  Regression and artificial neural network models for strength properties of engineered cementitious composites , 2016, Neural Computing and Applications.

[77]  D. Opitz,et al.  Popular Ensemble Methods: An Empirical Study , 1999, J. Artif. Intell. Res..

[78]  Nele De Belie,et al.  Self-Healing in Cementitious Materials—A Review , 2013 .

[79]  Victor C. Li,et al.  Rheology, fiber dispersion, and robust properties of Engineered Cementitious Composites , 2013 .

[80]  V. Li,et al.  Tailoring engineered cementitious composites for impact resistance , 2012 .

[81]  V. Li,et al.  Effect of Cracking and Healing on Durability of Engineered Cementitious Composites Under Marine Environment , 2007 .

[82]  Bart Verheij,et al.  Artificial intelligence as law , 2020, Artificial Intelligence and Law.

[83]  François-Xavier Coudert,et al.  Machine learning approaches for the prediction of materials properties , 2020, APL Materials.

[84]  V. Li.  Engineered Cementitious Composites (ECC): Bendable Concrete for Sustainable and Resilient Infrastructure , 2019 .

[85]  Zhongya Zhang,et al.  Artificial neural networks applied to polymer composites: a review , 2003 .

[86]  Karen L. Scrivener,et al.  Options for the future of cement , 2014 .

[87]  K. Hossain,et al.  Structural Performance of Polymer Fiber Reinforced Engineered Cementitious Composites Subjected to Static and Fatigue Flexural Loading , 2015 .

[88]  V. Li,et al.  TENSILE STRAIN-HARDENING BEHAVIOR OF POLYVINYL ALCOHOL ENGINEERED CEMENTITIOUS COMPOSITE (PVA-ECC) , 2001 .

[89]  V. Li.  On Engineered Cementitious Composites (ECC) , 2003 .

[90]  En-Hua Yang,et al.  Use of High Volumes of Fly Ash to Improve ECC Mechanical Properties and Material Greenness , 2007 .

[91]  G. Ma,et al.  Hybrid effects of steel fibers, basalt fibers and calcium sulfate on mechanical performance of PVA-ECC containing high-volume fly ash , 2019, Cement and Concrete Composites.

[92]  C. Leung,et al.  A fundamental method for prediction of failure of strain hardening cementitious composites without prior information , 2020 .

[93]  Shengyong Chen,et al.  Artificial Intelligence in Civil Engineering , 2012 .