F. Kamran, N. Jhora, Taher Baghaee Moghaddam, Leila Hashemian
Asphalt pavement is demolished and milled at the end of its service life, resulting in an enormous amount of waste material known as reclaimed asphalt pavement (RAP). Using cold recycling techniques, it is possible to reuse RAP for base course construction. However, using a high amount of RAP could have negative effects on the mechanical properties and durability of the base course. In this study, to improve the performance of asphalt emulsion recycled base course prepared with RAP, asphaltenes derived from Alberta oil-sands bitumen was added to the mixture. The performance properties of the selected mixtures were investigated at high, intermediate and low temperatures. Testing results and statistical analysis revealed that asphaltenes enhances the high and intermediate temperature performance of the modified mixes significantly. The low-temperature creep compliance of asphaltenes-modified mixtures was slightly lower than the unmodified mixtures, however, the low-temperature properties of the modified mixes were not significantly affected.
{"title":"Assessment of Performance Characteristics of Pavement Base Course Incorporating Reclaimed Asphalt Pavement and Asphaltenes","authors":"F. Kamran, N. Jhora, Taher Baghaee Moghaddam, Leila Hashemian","doi":"10.1139/cjce-2023-0419","DOIUrl":"https://doi.org/10.1139/cjce-2023-0419","url":null,"abstract":"Asphalt pavement is demolished and milled at the end of its service life, resulting in an enormous amount of waste material known as reclaimed asphalt pavement (RAP). Using cold recycling techniques, it is possible to reuse RAP for base course construction. However, using a high amount of RAP could have negative effects on the mechanical properties and durability of the base course. In this study, to improve the performance of asphalt emulsion recycled base course prepared with RAP, asphaltenes derived from Alberta oil-sands bitumen was added to the mixture. The performance properties of the selected mixtures were investigated at high, intermediate and low temperatures. Testing results and statistical analysis revealed that asphaltenes enhances the high and intermediate temperature performance of the modified mixes significantly. The low-temperature creep compliance of asphaltenes-modified mixtures was slightly lower than the unmodified mixtures, however, the low-temperature properties of the modified mixes were not significantly affected.","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140079795","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
B. Kadhom, Eli Simova, Jieying J Zhang, Dennis Krys
In this paper, the suitability of digital image correlation (DIC) technology in evaluating the flexural behaviour of as-built and carbon fiber reinforced polymer (CFRP) strengthened reinforced concrete (RC) beams was verified by conducting a 4-point bending test on four large-size beams. The DIC data obtained during the tests were compared to the results measured using traditional techniques such as displacement sensors and electrical strain gauges, as well as the findings derived from finite element (FE) numerical simulations. A good agreement was achieved between the local displacement and strain measurements and the data from DIC as a function of the applied load. Moreover, the crack patterns generated by the FE modelling were validated by the corresponding patterns derived from DIC. This suitability study is expected to contribute to the future field implementation of DIC technology to monitor the CFRP-strengthened members of critical structures such as bridges.
{"title":"Suitability of DIC in Capturing Flexural Behaviour of As-built and CFRP‑Strengthened RC Beams","authors":"B. Kadhom, Eli Simova, Jieying J Zhang, Dennis Krys","doi":"10.1139/cjce-2023-0154","DOIUrl":"https://doi.org/10.1139/cjce-2023-0154","url":null,"abstract":"In this paper, the suitability of digital image correlation (DIC) technology in evaluating the flexural behaviour of as-built and carbon fiber reinforced polymer (CFRP) strengthened reinforced concrete (RC) beams was verified by conducting a 4-point bending test on four large-size beams. The DIC data obtained during the tests were compared to the results measured using traditional techniques such as displacement sensors and electrical strain gauges, as well as the findings derived from finite element (FE) numerical simulations. A good agreement was achieved between the local displacement and strain measurements and the data from DIC as a function of the applied load. Moreover, the crack patterns generated by the FE modelling were validated by the corresponding patterns derived from DIC. This suitability study is expected to contribute to the future field implementation of DIC technology to monitor the CFRP-strengthened members of critical structures such as bridges.","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140090363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Engineered wood products along with advanced processing and fabrication are pushing the limits of modern wood construction. Innovative concepts and applications facilitate the development of a new generation of structural systems. However, capacities of such systems are often governed by capabilities of the connections. This paper presents the experimental results with numerical and analytical models of angle brackets for prediction of load bearing capacity, stiffness and ductility. Three types of metal brackets in beam-column connections have been investigated and their performance has been studied in various loading arrangements. Detailed finite-element models of each connection have been developed to gain insights into their behavior. An analytical approach is also adopted to represent the connections. Comparison with test data suggests that the models can reproduce results with good accuracy. The findings confirm feasibility of implementing ductile connections in practical mass timber structures.
{"title":"Experimental, Finite Element and Analytical Characterization of Hysteretic Response of Ductile Connections with Nailed Angle Brackets for Mass Timber Structures","authors":"P. Sejkot, A. Aloisio, Vera Obradovic, Asif Iqbal","doi":"10.1139/cjce-2023-0394","DOIUrl":"https://doi.org/10.1139/cjce-2023-0394","url":null,"abstract":"Engineered wood products along with advanced processing and fabrication are pushing the limits of modern wood construction. Innovative concepts and applications facilitate the development of a new generation of structural systems. However, capacities of such systems are often governed by capabilities of the connections. This paper presents the experimental results with numerical and analytical models of angle brackets for prediction of load bearing capacity, stiffness and ductility. Three types of metal brackets in beam-column connections have been investigated and their performance has been studied in various loading arrangements. Detailed finite-element models of each connection have been developed to gain insights into their behavior. An analytical approach is also adopted to represent the connections. Comparison with test data suggests that the models can reproduce results with good accuracy. The findings confirm feasibility of implementing ductile connections in practical mass timber structures.","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-02-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140410537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Wildfire impacts are increasing due to the multiplicative effect of several factors, including climate change, increased vulnerability in the wildland-urban interface, and impacts of management decisions. This has also led to an increase in evacuations due to the number of wildfires and people affected. This study collected information on behaviour during wildfire evacuation to fill critical research gaps in human behaviour and evacuation knowledge. Seven videos of residents’ evacuations from the 2016 Fort McMurray fire were collected from public platforms. Their routes were analyzed, and notable behavioural events were recorded. The evacuees mainly used major roads before getting onto the highway (the only route available for vehicular egress). The notable behaviours observed included using opposite lanes and driving outside of marked roads to avoid congestion. Much of the observed behaviours appeared to be motivated by the surrounding traffic or fire behaviour, further supporting the need for further studies of evacuation.
{"title":"Video Analysis of Human Behaviour during Wildfire Evacuations","authors":"Hannah Carton, John Gales, Eric B Kennedy","doi":"10.1139/cjce-2023-0450","DOIUrl":"https://doi.org/10.1139/cjce-2023-0450","url":null,"abstract":"Wildfire impacts are increasing due to the multiplicative effect of several factors, including climate change, increased vulnerability in the wildland-urban interface, and impacts of management decisions. This has also led to an increase in evacuations due to the number of wildfires and people affected. This study collected information on behaviour during wildfire evacuation to fill critical research gaps in human behaviour and evacuation knowledge. Seven videos of residents’ evacuations from the 2016 Fort McMurray fire were collected from public platforms. Their routes were analyzed, and notable behavioural events were recorded. The evacuees mainly used major roads before getting onto the highway (the only route available for vehicular egress). The notable behaviours observed included using opposite lanes and driving outside of marked roads to avoid congestion. Much of the observed behaviours appeared to be motivated by the surrounding traffic or fire behaviour, further supporting the need for further studies of evacuation.","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140429856","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Huanru Zhu, M. Bezabeh, Asif Iqbal, M. Popovski, Zhiyong Chen
Post-tensioned cross-laminated timber (PT-CLT) walls have been demonstrated to be a low-damage seismic force-resisting system (SFRS) due to their self-centring capability. However, there is still a need to examine the seismic performance of such SFRS in high seismic risk zones. This study evaluates the seismic performance of 6-, 9-, and 12-storey PT-CLT shear wall buildings in Vancouver, Canada, equipped with Buckling-Restrained Axial Fuses (BRAFs). The prototype buildings were designed using the displacement-based design method, and the assessment considered the most recent seismic hazard model provided in the 2020 National Building Code of Canada. To conduct nonlinear response history analysis (NLRHA) and incremental dynamic analysis (IDA), numerical models were developed in OpenSeesPy and calibrated based on component and system-level experimental tests. The NLRHA and IDA results demonstrate that all the studied buildings have adequate collapse margin ratios, with less than a 10% chance of collapsing at maximum considered earthquakes.
{"title":"Seismic Performance Assessment of Post-Tensioned CLT Shear Wall Buildings with Buckling-Restrained Axial Fuses","authors":"Huanru Zhu, M. Bezabeh, Asif Iqbal, M. Popovski, Zhiyong Chen","doi":"10.1139/cjce-2023-0448","DOIUrl":"https://doi.org/10.1139/cjce-2023-0448","url":null,"abstract":"Post-tensioned cross-laminated timber (PT-CLT) walls have been demonstrated to be a low-damage seismic force-resisting system (SFRS) due to their self-centring capability. However, there is still a need to examine the seismic performance of such SFRS in high seismic risk zones. This study evaluates the seismic performance of 6-, 9-, and 12-storey PT-CLT shear wall buildings in Vancouver, Canada, equipped with Buckling-Restrained Axial Fuses (BRAFs). The prototype buildings were designed using the displacement-based design method, and the assessment considered the most recent seismic hazard model provided in the 2020 National Building Code of Canada. To conduct nonlinear response history analysis (NLRHA) and incremental dynamic analysis (IDA), numerical models were developed in OpenSeesPy and calibrated based on component and system-level experimental tests. The NLRHA and IDA results demonstrate that all the studied buildings have adequate collapse margin ratios, with less than a 10% chance of collapsing at maximum considered earthquakes.","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140436642","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Capturing and modeling reality to enhance the built environment","authors":"Ala Suliman, Zheng Liu, Tarek Hegazy","doi":"10.1139/cjce-2024-0027","DOIUrl":"https://doi.org/10.1139/cjce-2024-0027","url":null,"abstract":"","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140440593","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xinhao Zhou, Lin Zhao, Zhaodong Liu, Liping Fu, Guangyuan Pan
Timely road surface condition (RSC) monitoring and maintenance significantly influences road safety. The current RSC relies on fixed road surveillance cameras and in-vehicle cameras. However, the fixed camera demands higher precision, while the in-vehicle camera requires higher timeliness. To address these challenges, this paper introduces an adaptive machine learning framework for simultaneous road surface detection on both device types. Initially, a convolutional neural network -based differentiation module identifies image sources. Subsequently, an adaptive algorithm switching mechanism leads to the development of two algorithms improved upon the real-time object detection algorithms. At last, extensive experiments with datasets collected from Ontario, Canada and Iowa U.S. validate the framework. Results show satisfactory classification accuracy, detection precision, and speed. Notably, the Mean Average Precision, namely mean of the average Precision for all categories(mAP)reaches 91.9% for fixed cameras and 90.6% for in-vehicle cameras, outperforming existing road surface snow detection models.
{"title":"An Adaptive Machine Learning Framework for Multi-Scenes Road Surface Weather Condition Monitoring","authors":"Xinhao Zhou, Lin Zhao, Zhaodong Liu, Liping Fu, Guangyuan Pan","doi":"10.1139/cjce-2023-0405","DOIUrl":"https://doi.org/10.1139/cjce-2023-0405","url":null,"abstract":"Timely road surface condition (RSC) monitoring and maintenance significantly influences road safety. The current RSC relies on fixed road surveillance cameras and in-vehicle cameras. However, the fixed camera demands higher precision, while the in-vehicle camera requires higher timeliness. To address these challenges, this paper introduces an adaptive machine learning framework for simultaneous road surface detection on both device types. Initially, a convolutional neural network -based differentiation module identifies image sources. Subsequently, an adaptive algorithm switching mechanism leads to the development of two algorithms improved upon the real-time object detection algorithms. At last, extensive experiments with datasets collected from Ontario, Canada and Iowa U.S. validate the framework. Results show satisfactory classification accuracy, detection precision, and speed. Notably, the Mean Average Precision, namely mean of the average Precision for all categories(mAP)reaches 91.9% for fixed cameras and 90.6% for in-vehicle cameras, outperforming existing road surface snow detection models.","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140443797","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Marcos V. G. Silveira, Juliana S. Wagner, Mohsen Khanverdi, Sreekanta Das
Despite the growing incorporation of additive manufacturing in the construction industry, a significant gap persists in the availability of design guidelines and comprehensive structural assessments for 3D-printed components. This research employs experimental methods to evaluate the performance of large-scale hollow unreinforced 3D-printed masonry walls under axial compressive load. Furthermore, a design methodology tailored for 3D-printed masonry walls under compressive loads is proposed in compliance with both American and Canadian masonry standards. This methodology is supported by the application of linear elastic 3D finite element analysis. Moreover, supplementary insights into material behavior are garnered through the examination of small specimens extracted from an additional 3D-printed wall. The results reveal that 3D printing reduces stiffness and compressive strength. The 3D-printed infills played an important role in redistributing stress after cracking. All full-scale 3D-printed masonry walls examined in this research exceeded the performance criteria set by the American masonry standard for load-bearing walls.
尽管增材制造技术越来越多地应用于建筑行业,但在三维打印构件的设计指南和全面结构评估方面仍存在巨大差距。本研究采用实验方法评估了大型空心非加固 3D 打印砌体墙在轴向压缩荷载作用下的性能。此外,还根据美国和加拿大的砌体标准,提出了在压缩荷载作用下为 3D 打印砌体墙量身定制的设计方法。线性弹性三维有限元分析的应用为该方法提供了支持。此外,还通过对从另外一个 3D 打印墙体中提取的小试样进行检查,获得了对材料行为的补充见解。结果显示,三维打印降低了刚度和抗压强度。三维打印填充物在开裂后重新分配应力方面发挥了重要作用。本研究中检测的所有全尺寸 3D 打印砌体墙均超过了美国砌体承重墙标准规定的性能标准。
{"title":"Structural Performance of Large-Scale 3D-Printed Walls Subjected to Axial Compression Load","authors":"Marcos V. G. Silveira, Juliana S. Wagner, Mohsen Khanverdi, Sreekanta Das","doi":"10.1139/cjce-2023-0395","DOIUrl":"https://doi.org/10.1139/cjce-2023-0395","url":null,"abstract":"Despite the growing incorporation of additive manufacturing in the construction industry, a significant gap persists in the availability of design guidelines and comprehensive structural assessments for 3D-printed components. This research employs experimental methods to evaluate the performance of large-scale hollow unreinforced 3D-printed masonry walls under axial compressive load. Furthermore, a design methodology tailored for 3D-printed masonry walls under compressive loads is proposed in compliance with both American and Canadian masonry standards. This methodology is supported by the application of linear elastic 3D finite element analysis. Moreover, supplementary insights into material behavior are garnered through the examination of small specimens extracted from an additional 3D-printed wall. The results reveal that 3D printing reduces stiffness and compressive strength. The 3D-printed infills played an important role in redistributing stress after cracking. All full-scale 3D-printed masonry walls examined in this research exceeded the performance criteria set by the American masonry standard for load-bearing walls.","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140442080","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper introduces a new low-damage hybrid steel–timber rocking braced frame with self-centring capability and proposes a design methodology in the framework of Canadian design standards. This system consists of a gravity load-resisting timber frame and a steel concentrically braced frame acting as the lateral load-resisting system that is connected using U-shaped steel fuses. The self-centring capability is achieved by post-tensioning strands and rocking bases within the lateral load-resisting system. The system, along with its components, is first introduced. The proposed design method and system performance objectives are then presented. The seismic performance of the system and the proposed design procedure are validated through numerical analyses considering the interaction between the lateral and gravity load-resisting systems. The results of dynamic analyses confirm the enhanced seismic performance of the proposed hybrid system with almost no residual deformations after the earthquake. Moreover, the proposed design guidelines can accurately predict the system’s seismic response and demands imposed on the members.
本文介绍了一种具有自聚能的新型低损坏混合钢木摇动支撑框架,并在加拿大设计标准框架内提出了一种设计方法。该系统由重力承重木框架和作为侧向承重系统的钢同心支撑框架组成,两者采用 U 型钢引信连接。抗侧荷载系统中的后张力钢绞线和摇动基座实现了自集中能力。首先介绍了该系统及其组成部分。然后介绍了拟议的设计方法和系统性能目标。考虑到抗侧力和重力荷载系统之间的相互作用,通过数值分析验证了系统的抗震性能和建议的设计程序。动态分析结果证实了所提出的混合系统的抗震性能得到了增强,地震后几乎没有残余变形。此外,所提出的设计准则能够准确预测系统的地震响应和对构件的要求。
{"title":"A new hybrid steel–timber rocking seismic force resisting system equipped with U-shaped fuse connections","authors":"Ahmed Mowafy, A. Imanpour, Ying-Hei Chui","doi":"10.1139/cjce-2023-0319","DOIUrl":"https://doi.org/10.1139/cjce-2023-0319","url":null,"abstract":"This paper introduces a new low-damage hybrid steel–timber rocking braced frame with self-centring capability and proposes a design methodology in the framework of Canadian design standards. This system consists of a gravity load-resisting timber frame and a steel concentrically braced frame acting as the lateral load-resisting system that is connected using U-shaped steel fuses. The self-centring capability is achieved by post-tensioning strands and rocking bases within the lateral load-resisting system. The system, along with its components, is first introduced. The proposed design method and system performance objectives are then presented. The seismic performance of the system and the proposed design procedure are validated through numerical analyses considering the interaction between the lateral and gravity load-resisting systems. The results of dynamic analyses confirm the enhanced seismic performance of the proposed hybrid system with almost no residual deformations after the earthquake. Moreover, the proposed design guidelines can accurately predict the system’s seismic response and demands imposed on the members.","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140443833","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study investigates the utility of machine learning (ML) in understanding and mitigating winter road risks. Despite their capability in managing complex data structures, ML models often lack interpretability. We address this issue by integrating Shapley Additive Explanations (SHAP) with a Support Vector Machine (SVM) model. Utilizing a comprehensive dataset of 231 snowstorm events collected in the city of Edmonton across two winter seasons, the SVM model achieved an accuracy rate of 87.2%. Following model development, a SHAP summary plot was employed to identify the contribution of individual features to collision predictions—an insight not achievable through ML alone. Next, SHAP waterfall plots were used to assess the reliability of individual predictions. The findings enhanced our understanding of the complex SVM model and provided greater insights into the diverse factors affecting winter road safety.
{"title":"Using Explainable AI for Enhanced Understanding of Winter Road Safety: Insights with Support Vector Machines and SHAP","authors":"Zehua Shuai, Tae J. Kwon, Qian Xie","doi":"10.1139/cjce-2023-0446","DOIUrl":"https://doi.org/10.1139/cjce-2023-0446","url":null,"abstract":"This study investigates the utility of machine learning (ML) in understanding and mitigating winter road risks. Despite their capability in managing complex data structures, ML models often lack interpretability. We address this issue by integrating Shapley Additive Explanations (SHAP) with a Support Vector Machine (SVM) model. Utilizing a comprehensive dataset of 231 snowstorm events collected in the city of Edmonton across two winter seasons, the SVM model achieved an accuracy rate of 87.2%. Following model development, a SHAP summary plot was employed to identify the contribution of individual features to collision predictions—an insight not achievable through ML alone. Next, SHAP waterfall plots were used to assess the reliability of individual predictions. The findings enhanced our understanding of the complex SVM model and provided greater insights into the diverse factors affecting winter road safety.","PeriodicalId":9414,"journal":{"name":"Canadian Journal of Civil Engineering","volume":null,"pages":null},"PeriodicalIF":1.4,"publicationDate":"2024-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140447153","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}