下载排行
Application of knowledge graph and representation learning in the mining of road traffic accident data
YU Dexin;PENG Wanli;WU Xincheng;CHEN Yunjie;LIU Xiaojia;The volume of data in traffic safety is substantial, containing rich semantic information. Extracting value from extensive road traffic accident data can provide crucial knowledge support for accident prevention and informed decision-making. However, conventional accident analysis methods face limitations when handling complex and multi-semantic intersecting accident data. Hence, this study proposes a method for mining accident data based on knowledge graph and knowledge representation learning. This method concatenates accident information by leveraging the graph structure and extracts associated information in spatial vectors, thereby enhancing the likelihood of discovering potential value in traffic accident data mining tasks. We establish a knowledge structure schema aligned with the general logic structure of road traffic accident occurrence and then proceed to construct a road traffic accident knowledge graph. In the mining process, we employ the TransE model for representation learning, wherein accident entities and causal relationships are mapped into vector space. Subsequently, semantic information between vectors is captured through vector matching operations to extract potential traffic accident information. In our initial experiments, we compare the model's performance on various datasets and vector dimensions, and we assess our method with real road accident data in different prediction tasks. The experimental results indicate that the proposed method exhibits high accuracy and robust semantic parsing capabilities. Additionally, utilizing vector spaces with appropriate dimensions enhances the accuracy of predictions for traffic-related data. This approach enables effective extraction of potential value from road traffic accident data. In fact, it helps mitigate biases caused by missing accident data in accident analysis and prevention decision-making processes, thereby enhancing the overall completeness of road traffic accident data. Therefore, the research results propose a novel method and idea for maximizing the utilization of fragmented information on road traffic accidents.
Safety risk assessment and influence degree analysis of the flight area based on the matter-element extension theory
FU Ning;SHAO Yueling;To estimate the safety level of airport flight area systems in operation effectively and security, a model for assessing safety level based on the matter-element extension theory model is constructed. The airport flight area is characterized by numerous and complex risk factors, which are prone to aviation accidents. Scientific and effective assessment of airport flight area safety risks has significant implications for the improvement of operation safety. This paper is divided into four parts. In the first part, we divide the influencing factors into four categories, a total of 24 risk factors based on the “man-machine-environment-management” system engineering theory. In the second part, we construct a multi-level assessment index system for the safety risk of the airport flight area. In the third part, the DEMATEL-ISM method is used to analyze the interaction degree of risk factors. We build a multi-level hierarchical structure model to express the complex relationship and analyze the interaction between these risk factors by finding out the surface, middle and deep factors. In the fourth part, considering that the determination of a single weight is susceptible to subjective factors and ignores the correlation between risk factors. A new combination weighting approach is proposed using DEMATEL and AHP methods. And matter-element extension evaluation model is established based on game theory combination weighting to evaluate the safety risk of the airport flight area. Besides, we use a case to show this model. The evaluation results show that the risk level of the airport is medium and the safety awareness of the airport staff needs to be strengthened. The viability and confidence of the model in the safety evaluation of the flight area in operation are proved. This safety assessment has a guiding significance to the flight area system and provides a base for further research.
Study on evaluation models of severity degree of dam failure impact
LI Lei, WANG Ren-zhong, SHENG Jin-bao(Dam safety and management department of Nanjing Hydraulic Research Institute, Dam Safety Management Center of the Ministry of Water Resources, Nanjing 210029, China)Dam risk assessment technique has already been used in the field of dam safety management abroad especially in Canada, Australia and USA in recent twenty years. There are more than 85 000 dams in China and about one-third of them are with severe deficiency and in high risks. Therefore it is real necessary to use this kind of technique to assist prioritizing dam maintenance or rehabilitation, reducing risks caused by dam failure. The technique consists of two important research objectives, dam-break probability analysis and dam failure impact evaluation. Dam failure impact includes three main factors: loss of life, financial loss, society and environment. How to evaluate the severity degree by means of these three factors comprehensively is a challenge of dam rehabilitation decision-making. The paper is to present a comprehensive evaluation coefficient for dam-break severity degree. The study introduces integrative assessment function L which considered weights Si respectively of life loss, financial loss, society and environment, and their severity coefficients Fi, formed as linear weighted sum method. In order to integrate loss of life, financial loss, society and environment in accordance with existing laws and regulations in China, logarithm non-linear or linear models of data normalization are established to deal with their different units. A series of quantitative values of L which divide disaster event into deferent degree are suggested according to Chinese situation. The method and evaluation model is practically applied at 5 reservoir dams to appraise their severity degree of failure impact respectively and the comparison of these results are made to decide which one is more severe. The analyzed result shows that hazard of all these 5 dams would be extreme severe and should report to State Council as soon as the event of dam break occur.
Research on personalized safety management for construction workers considering personality traits
HU Zhe;HU Hao;XU Feng;WANG Wen;This paper intends to investigate the role of individual factors(personality traits and injury experience) on construction workers' unsafe behaviors and find out a way to incorporate them into improving the effectiveness of current on-site safety management. For the given purpose, three steps have been conducted. Firstly, the hypothesis on the relationship between injury experience and risk propensity was made based on the cognitive psychology and big-five theory, as well as the moderating effects of personality traits(OCEAN) on the route of risk propensity leading to unsafe acts. Secondly, the method of constructing a discriminant tool using personality traits to identify the level of unsafe behavior intentions was studied. Lastly, the results of the analysis were applied to design a personalized safety management program involving safety instruction and safety training. 533 valid data on injury experience, personality traits, risk propensity, and unsafe behavior intention were found in the previous social survey. The data were grouped in terms of risk propensity and behavior intention based on the cumulative probability distribution. Through the ANOVA analysis, it has been found that there is a significant difference in risk propensity between the workers with injury experience and the workers without injury experience. The worker with a high-risk propensity or unsafe behavior intention has been found to have the common characteristics of low agreeableness and conscientiousness, and high openness and extraversion. Openness, extraversion, and agreeableness were found to be mediators in the relationship between risk propensity and unsafe behavior intention through the moderating analysis. Moreover, the former two have positive effects and the latter has a negative effect. Besides, it has been found that the unsafe behavior intention can be identified by personality traits through the discriminant analysis. Thus, it can be concluded that personality traits are the indicators of unsafe behavior intention and should be considered in safety practice. With the help of technologies such as IoT and wearable sensors, the contents and delivery method design of safety instruction can take personality traits into account, as well as the contents and intensity setting of safety training, to obtain the personalization of construction safety management.
Statistics of industrial accidents in China during the period from September to October in 2021
LI Sheng-cai;XIAO Lei;The present paper reports the statistical data and analysis of industrial accidents in China during the period from September to October in 2021. There are 63 accidents in total, which can be summarized and classified into six categories: traffic, mining industry, chemical industry, architecture engineering, municipal engineering and the others. Of all the 63 accidents, the percentages of the six categories can be illustrated according to their frequency order 38.10%, 6.35%, 12.70%, 3.17%, 25.40%, and 14.29% respectively. The death toll is 195 and the percentages for the six categories are 48.72%, 5.64%, 11.28%, 1.03%, 21.03%, and 12.31% respectively. The analyses of the statistics show that the top five provinces and autonomous regions with highest accident number are: Zhejiang(5 accidents), Sichuan(5), Inner Mongolia(5), Heilongjiang(5), and Liaoning(4). The top five provinces and autonomous regions with the largest death toll are: Heilongjiang(25 dead), Hebei(25), Anhui(20), Inner Mongolia(15), and Liaoning(14).
Vibration characteristic analysis of double-elbow natural gas pipeline with fluid-solid interaction
ZHANG Peng;YAO Zheng-xue;LIU Si-ming;To avoid the resonance of gas column in long-distance pressured transportation of natural gas, the paper uses the transfer matrix method to calculate the natural frequency of the air column, and compares them with the results of the finite element calculation. The results obtained by the two methods are almost identical with small errors. It shows that acoustic simulation can more accurately calculate the frequency of the air column in the simple pipe. Hereafter we analyzed the modal characteristics of the pipeline. The low-order mode of the double-elbow pipeline that transports natural gas is dominated by translational motion in the X and Y directions, and no torsional vibration occurs. The fluid-solid interaction does not affect the vibration shape of the pipeline, but it can reduce the natural frequency of the pipeline. The gas-solid interaction reduces the natural frequency of the double-elbow pipeline by about 2%, and the liquid-solid interaction reduces the natural frequency of the pipeline by about 14.5%, so the fluid-solid interaction should be considered in the modal analysis. Subsequently, the influence of the pulsation amplitude and frequency of the fluid on the vibration of the pipeline was studied. The pipe amplitude increases with the increase of the fluid pulsation amplitude. When the fluid pulsation amplitude is close to or equal to the natural frequency of the pipeline, resonance will occur, and the pipeline amplitude will become larger and larger in a short time. Therefore, special attention should be paid to fluid conditions during transportation. Furthermore, The inner diameter of the pipe and the radius of curvature of the elbow also have an important influence on the vibration response of the pipe. In the pipeline design, the inner diameter of the pipeline should be increased reasonably according to the flow conditions, and the long radius elbow should be used to reduce the exciting force.
Safety risk assessment of highway slope construction based on multidimensional contact cloud model
CHEN Zhao;SUN Jingnan;ZHOU Zilong;SHI Zhihai;To better avoid the losses caused by highway slope construction accidents in mountainous areas and obtain hierarchical management and effective prevention and control of risks during slope construction, the safety risk assessment index system and risk level standard of highway slope construction in mountainous areas are established according to the actual risk identification of a highway slope construction in Guangxi Province and the current risk assessment index of slope construction. Firstly, the subjective and objective weights of each slope construction risk assessment index are respectively determined by the G1 method and the improved CRITIC(Criteria Importance Though Intercrieria Correlation) method and then optimized by using the game theory to obtain the optimal weight of each risk assessment index. Secondly, by adopting the same different opposite principle of set pair analysis and the 3En criterion of the cloud model, the cloud connection degree of the multidimensional cloud model is improved. A multidimensional connection cloud model that can take into account the uncertainty, randomness, and fuzziness of evaluation indexes and reflect interactions between evaluation indexes is established. Finally, the established model is applied to the construction safety risk assessment of six highway slope construction cases in Guangxi Province, and the construction risk levels of six slopes are determined. Meanwhile, the determined construction risk level of each slope is compared to the evaluation results based on the multidimensional cloud model, set pair cloud model, cloud model, set pair analysis method, and index system method. Results indicate that the multi-dimensional connection cloud model not only ensures the consistency of evaluation results but also realizes the determination of slope safety risk transformation trends in the construction process. Therefore, the established safety risk assessment index system and risk level standard of highway slope construction in mountainous areas are effective and reliable, which is more conducive to risk management and effective prevention and control of slope construction.
Co-citation analysis of the Chinese international safety literatures from 2018to 2020:intellectual bases and research directions
QIAN Jin-xin;LI Sheng-cai;GAN Qiang;LI Jie;To understand the intellectual bases and structure of Chinese safety science research from 2018 to 2020,and put forward significant guides to the development of Chinese Safety Science,2 510 publications(articles and reviews) from 23 safetyrelated journals have been downloaded from the Web of Science core collection-Science Citation Index.The co-citation method was applied to measure the similarities of the cited authors,cited journals,and cited references.And the scientometrics tools,including BAT,VOS viewer and CiteSpace were used in the data pre-processing and co-citation network visualization.The data was analyzed from three dimensions,the journal level,the author level,and the reference level.The journal and author networks and clusters were obtained by VOS viewer,and the references co-citation network and clusters were generated by CiteSpace.The analysis results show that high cited journals,including Reliab Eng Syst Safe,Accident Anal Prev,Safety Sci,and J Loss Prevent Proc are core knowledge sources in Chinese Safety research.The co-citation network of journals was clustered into six groups,including the 'system safety and reliability analysis 'disasters risk' 'environment risk' 'safety science and ergonomics process safety ',and traffic safety '.At the author level,Levitin G(Israel Elect Corp Ltd),Khakzad N(Ryerson Univ),and Khan FI(Mem Univ Newfoundland) are highly cited authors by Chinese Safety community,reflecting that these authors' research topics are highly relevant to Chinese Safety research in recent years.The clusters of the author co-citation network are mainly from four aspects 'safety management science' 'reliability and system safety' 'process safety',and'traffic safety'.At the reference level,the papers from Levitin G(2005),Reason J(1990) and Leveson N(2004),etc.are highly cited by Chinese researchers,and these highly cited papers are highly cited in 'domino effect' 'mission abort policy ''degradation-analysis' etc.,from Chinese safety science community.
Experiment and molecular simulation study on the adsorption of CH4 and CO2 by coal with high-metamorphic
QU Guo-na;MA Zhao-xin;JIA Ting-gui;LOU He-zhuang;QIANG Qian;HAO Yu;To explore the adsorption mechanism of anthracite on CH4 and CO2 gas, this paper mainly selected the high metamorphic anthracite from No.2 well of the Sihe Coal Mine in Jincheng of Shanxi. Then the adsorption characteristics of single component CH4 and CO2 gases were tested by using the high-pressure capacity method gas adsorption equipment. The adsorption behavior that occurred in the established chemical structure model of anthracite molecules was simulated through the Grand Canonical Monte Carlo(GCMC) to analyze the properties of anthracite for adsorbing CH4 and CO2 gases. The effects of anthracite on adsorbing CH4 and CO2 gases under different temperature conditions are discussed, and the relation between the adsorption capacity of CH4 and CO2 and the adsorption heat of CH4 and CO2 is fitted. The results show that, under the isothermal condition, the adsorption capacity of anthracite for CH4 and CO2 gas increases with the increase of pressure. On the contrary, the growth momentum of adsorption capacity slows down gradually. Under constant pressure, the adsorption capacity of anthracite for CH4 and CO2 gases decreases gradually with the increase of temperature. CO2 molecules with stronger adsorption capacity displace CH4 molecules with weaker adsorption capacity from the coal surface. Consequently, the adsorption capacity of anthracite for CO2 gas is greater than CH4 gas. Moreover, the adsorption heat of anthracite for CH4 gas increases with increasing the adsorption capacity, showing a positive correlation. On the other hand, the adsorption heat of anthracite for CO2 gas decreases with the increase of the adsorption capacity, showing a negative correlation.