Transferability of accident prediction models

Accident prediction models, the vast majority of which are negative binomial regression models, are of considerable importance to highway agencies since they can be used to conduct many traffic safety studies. However, not every agency possesses sufficient accident statistics that enable it to develop reliable models of its own. This problem gives rise to interest in the transferability of accident prediction models in time and space. It would save time, effort, and money if accident prediction models developed for one region in one period of time could be applied in different time periods and regions to produce reliable safety studies.This paper presents methods for recalibrating negative binomial accident models before transferring them for use in different time periods and regions of space. The paper emphasizes that the recalibration of the shape parameter of a transferred model using local data is absolutely necessary. It explains that it is also desirable to recalibrate the constant term of the transferred model in order to allow the model to better suit local conditions. A moment method is presented for recalibrating the shape parameter of a transferred model when its constant term is not recalibrated. However, a maximum likelihood method is presented for recalibrating both the shape parameter and the constant term of the transferred model and is shown to be superior to the recalibration methods existing in the traffic safety literature.     

基于链式多重插补的WOA-ELM煤与瓦斯突出预测模型

为了提高缺失数据下煤与瓦斯突出预测准确率,提出1种基于链式支持向量机多重插补(MICE＿SVM)的鲸鱼优化算法(WOA)-极限学习机(ELM)预测模型,以淮南朱集矿区为例,选取5个煤与瓦斯突出影响指标作为模型特征,采用提出的MICE＿SVM算法插补突出事故数据中缺失值,利用WOA优选ELM输入层权值及隐含层阈值,构建煤与瓦斯突出预测模型,将插补后数据用于WOA-ELM模型的训练与测试,并与其他模型的预测效果对比。研究结果表明：MICE＿SVM插补前、后的有突出数据预测准确率分别为83.02%,90.41%,MICE＿SVM显著提高了有突出预测准确率,对无突出和整体的预测准确率提高不明显;数据插补后WOA优化ELM对无突出、有突出和整体的预测准确率分别为97.94%,96.25%,96.48%,较优化前分别提高了5.79%,5.84%,5.55%,数据插补后WOA-ELM为最佳预测模型。     

Numerical study of medium to large scale BLEVE for blast wave prediction

So far, the prediction of blast wave generated from the Boiling Liquid Expanding Vapour Explosion (BLEVE) has been already broadly investigated. However, only a few validations of these blast wave prediction models have been made, and some well-established methods are available to predict BLEVE overpressure in the open space only. This paper presents numerical study on the estimation of the near-field and far-field blast waves from BLEVEs. The scale effect is taken into account by conducting two different scale BLEVE simulations. The expansion of pressurized vapour and evaporation of liquid in BLEVE are both modelled by using CFD method. Two approaches are proposed to determine the initial pressure of BLEVE source. The vapour evaporation and liquid flashing are simulated separately in these two approaches. Satisfactory agreement between the CFD simulation results and experimental data is achieved. With the validated CFD model, the results predicted by the proposed approaches can be used to predict explosion loads for better assessment of explosion effects on structures.     

基于GM(1,1)的残差修正模型的电梯故障率预测

为研究某城市某品牌电梯故障率发展趋势,建立了该城市该品牌电梯故障率的GM(1,1)灰色预测模型,并对所建模型进行了数据检验,检验结果表明该预测模型的预测精度波动较大。为了提高GM(1,1)灰色预测模型的预测精度,利用对模型进行数据检验时得到的残差序列,建立GM(1,1)灰色预测模型的残差修正模型,利用该残差修正模型对原预测模型进行修正。利用经残差修正模型修正后的故障率预测模型对该城市A品牌电梯的故障率进行预测,结果表明:1)残差修正模型对原模型修正后的相对误差与修正前相比有升也有降,但精度有所提高且趋于稳定,表明残差修正模型有利于提高预测精度;2)利用所建立的故障率预测模型求得的预测故障率与实际故障率相比,相对误差不超过8.010%,表明该故障率预测模型的预测精度较高;3)修正模型预测值表明,在现有状态下该城市A品牌电梯的故障率呈上升趋势,应加强该品牌电梯的检维修与管理。     

大气扩散模型统计性确认评估

使用计算模型预测物理过程之前,必须对该计算模型进行论证,确认评估通过后才能用于物理状态和过程的计算和预测。讨论了基于数据统计属性的确认评估对面向大气扩散物理现象的计算模型之适用性;探究了模型评估工作的基本内容、关键概念和通用原则;给出了确认评估参数的定义、选择原则及示例;分析了评估度量指标的内涵和适用条件;探讨了可接受模型需要满足的指标准则及统计性确认评估指标的选用;最后,指出了基于统计度量的确认评估方法的改进方向。扩散模型评估理论的探讨有助于提高基于模型仿真的大气扩散研究的准确性,也有利于高精度模型和基准试验的设计、开发与遴选。     

Using computational fluid dynamics (CFD) for blast wave predictions

Explosions will, in most cases, generate blast waves. While simple models (e.g., Multi Energy Method) are useful for simple explosion geometries, most practical explosions are far from trivial and require detailed analyses. For a reliable estimate of the blast from a gas explosion it is necessary to know the explosion strength. The source explosion may not be symmetric; the pressure waves will be reflected or deflected when hitting objects, or even worse, the blast waves may propagate inside buildings or tunnels with a very low rate of decay. The use of computational fluid dynamics (CFD) explosion models for near and far field blast wave predictions has many advantages. These include more precise estimates of the energy and resulting pressure of the blast wave, as well as the ability to evaluate non-symmetrical effects caused by realistic geometries, gas cloud variations and ignition locations. This is essential when evaluating the likelihood of a given leak source as cause of an explosion or equally when evaluating the potential risk associated with a given leak source for a consequence analysis.In addition, unlike simple methods, CFD explosion models can also evaluate detailed dynamic effects in the near and far field, which include time dependent pressure loads as well as reflection and focusing of the blast waves. This is particularly valuable when assessing actual near-field blast damage during an explosion investigation or potential near-field damage during a risk analysis for a facility. One main challenge in applying CFD, however, is that these models require more information about the actual facility, including geometry details and process information. Collecting the necessary geometry and process data may be quite time consuming. This paper will show some blast prediction validation examples for the CFD model FLACS. It will also provide examples of how directional effects or interaction with objects can significantly influence the dynamics of the blast wave. Finally, the challenge of obtaining useful predictions with insufficient details regarding the geometry will also be addressed.     

动态瓦斯地质预测算法研究与可视化应用

为实现煤矿瓦斯地质动态精准预测与可视化,基于煤层埋深、厚度、倾角、地质构造等多种因素,通过将相似的地质模块定义为同一微单元,在此基础上将反距离权重插值算法和递归邻域搜索策略与数据优化处理算法相结合,应用于瓦斯地质动态预测可视化系统开发。研究结果表明:根据递归邻域搜索算法模型开发的多级瓦斯地质图动态分析系统利用瓦斯地质基础数据可对矿井瓦斯含量、压力、涌出量等瓦斯赋存参数进行实时分析和计算,动态绘制散点图、等值线和区域预测图,使瓦斯区域预测图过渡更加平缓,显著提高预测的精准性及整体的显示效果,同时其制图耗时也无明显增加,可为高瓦斯矿井安全生产提供决策依据。     

CFD-based simulation of dense gas dispersion in presence of obstacles

Quantification of spatial and temporal concentration profiles of vapor clouds resulting from accidental loss of containment of toxic and/or flammable substances is of great importance as correct prediction of spatial and temporal profiles can not only help in designing mitigation/prevention equipment such as gas detection alarms and shutdown procedures but also help decide on modifications that may help prevent any escalation of the event.The most commonly used models - SLAB (Ermak, 1990), HEGADAS (Colenbrander, 1980), DEGADIS (Spicer & Havens, 1989), HGSYSTEM (Witlox & McFarlane, 1994), PHAST (DNV, 2007), ALOHA (EPA & NOAA, 2007), SCIPUFF (Sykes, Parker, Henn, & Chowdhury, 2007), TRACE (SAFER Systems, 2009), etc. - for simulation of dense gas dispersion consider the dispersion over a flat featureless plain and are unable to consider the effect of presence of obstacles in the path of dispersing medium. In this context, computational fluid dynamics (CFD) has been recognized as a potent tool for realistic estimation of consequence of accidental loss of containment because of its ability to take into account the effect of complex terrain and obstacles present in the path of dispersing fluid.The key to a successful application of CFD in dispersion simulation lies in the accuracy with which the effect of turbulence generated due to the presence of obstacles is assessed. Hence a correct choice of the most appropriate turbulence model is crucial to a successful implementation of CFD in the modeling and simulation of dispersion of toxic and/or flammable substances.In this paper an attempt has been made to employ CFD in the assessment of heavy gas dispersion in presence of obstacles. For this purpose several turbulence models were studied for simulating the experiments conducted earlier by Health and Safety Executive, (HSE) U.K. at Thorney Island, USA (Lees, 2005). From the various experiments done at that time, the findings of Trial 26 have been used by us to see which turbulence model enables the best fit of the CFD simulation with the actual findings. It is found that the realizable k-? model was the most apt and enabled the closest prediction of the actual findings in terms of spatial and temporal concentration profiles. It was also able to capture the phenomenon of gravity slumping associated with dense gas dispersion.     

基于不同指标无量纲化方法的岩爆预测模型优选

为研究不同指标无量纲化方法对岩爆等级预测模型精度的影响,提高岩爆预测准确率,选取应力系数、脆性系数和弹性能量指数作为预测指标。基于104组岩爆实例大样本数据,采用统一极差处理法、差异化极差处理法、平均化处理法和归一化处理法4种指标无量钢化方法,对预测指标的原始数据进行处理,建立不同的岩爆预测距离判别模型并进行工程实例应用。研究结果表明:基于平均化处理法的岩爆预测模型的回判准确率高达97.1%;对不同矿山、隧道和水电站的6个工程实例的预测结果符合实际情况,说明其是一种准确率高、方便实用的岩爆预测模型。     

单变量灰色预测模型在煤矿开采沉降预测中的对比分析

以预测煤矿开采而引起的地表高程的损失为目的,通过灰色系统理论的建模、关联度分析和残差辨识,建立基于贫信息的传统GM(1,1)模型、GM(1,1)残差模型、时序残差GM(1,1)模型,又建立基于原始数据具有绝对误差的灰色CompertzⅠ和灰色LogisticⅠ模型与具有相对误差的灰色CompertzⅡ和灰色LogisticⅡ模型,并将其应用到金竹山矿业公司土珠煤矿的地表沉降量的实际预测分析中,对该矿2007年度1—10月的地表高程损失量进行灰色生成后,建立了7种灰色预测模型。根据其预测值的精度检验结果对比分析表明,所建立的7种模型均为一级(好)模型,且灰色CompertzIⅡ和灰色LogisticIⅡ模型远优于传统GM(1,1)模型,预测精度高,可靠性强,对煤矿开采的复垦规划有重要指导作用。     

A beetle antennae search improved BP neural network model for predicting multi-factor-based gas explosion pressures

A gas explosion in an underground structure may cause serious damage to the human body and ground buildings and may result in huge economic losses. The pressure of the gas explosion is an important parameter in determining its severity and designating an emergency plan. However, existing empirical and computational fluid dynamics (CFD) methods for pressure prediction are either inaccurate or inefficient when considering multiple influencing factors and their interrelationships. Therefore, for a more efficient and reliable prediction, the present study developed a multifactorial prediction model based on a beetle antennae search (BAS) algorithm improved back propagation (BP) neural network. A total of 317 sets of data which considered factors of geometry, gas, obstacle, vent, and ignition were collected from previous studies. The results showed that the established model can predict pressures accurately by low RMSE (43.4542 and 50.7176) and MAPE (3.9666% and 4.9605%) values and high R² (0.7696 and 0.7388) values for training and testing datasets, respectively. Meanwhile, the BAS algorithm was applied to improve both the calculation efficiency and the accuracy of the proposed model by enabling a more intelligent hyperparameter tuning method. Furthermore, the permutation importance of input variables was investigated, and the length (L) and the ratio of length and diameter (L/D) of geometry were found to be the most critical factors that affect the explosion pressure level.     

国外灾害风险评估模型对比分析

灾害风险评估模型是灾害风险管理的科学工具,为灾害应急管理决策提供技术支持。本文对国外几个应用较为广泛的灾害风险评估模型进行了系统总结,分析了风险评估的基本流程、方法及要点,从评估方法、应用特点、使用效果等多个角度对这些模型的优缺点进行详细的对比分析。同时,总结了灾害风险评估模型具备良好的科学性和实用性应符合的基本原则,并以本文介绍的风险评估模型为例,对其基本原则的符合性进行了对比分析。分析结果表明,灾害风险评估模型只有在科学性与实用性之间获得最佳平衡点,并与应急规划、风险管理与减灾过程充分结合才能获得最佳的应用效果;建立灾害风险评估模型尤其要重视公众充分参与、在风险相关部门和人群中通过风险沟通建立统一的认识、准确界定脆弱性人群、有效解决数学模型或专家判断的不确定性等关键问题。     

基于支持向量回归机的煤层瓦斯含量预测研究

为了对煤层瓦斯含量进行准确预测,应用支持向量回归机(SVR)理论建立煤层瓦斯含量预测模型,结合现场实测数据利用支持向量机(SVM)工具箱进行模型的求解及预测,并从均方根误差、希尔不等系数和平均绝对百分误差3个不同误差指标与人工神经网络预测模型进行比较分析。研究结果表明:SVR模型其预测精度及可行性高于神经网络模型,而且运算快,实时性较好,用于煤层瓦斯含量的预测较理想,具有良好的应用前景,可以为煤矿瓦斯防治提供理论依据。     

Quantitative risk analysis of toxic gas release caused poisoning—A CFD and dose–response model combined approach

Some major toxic gas release accidents demonstrate the urgent need of a systematic risk analysis method for individuals exposed to toxic gases. A CFD numerical simulation and dose–response model combined approach has been proposed for quantitative analysis of acute toxic gas exposure threats. This method contains four steps: firstly, set up a CFD model and monitor points; secondly, solve CFD equations and predict the real-time concentration field of toxic gas releases and dispersions; thirdly, calculate the toxic dose according to gas concentration and exposure time; lastly, estimate expected fatalities using dose–response model. A case study of hydrogen sulfide releases from a gas gathering station has been carried out using a three dimension FLUENT model. Acute exposure fatalities have been evaluated firstly with a simplified ideal model which assumes workers stay at original exposure location without moving. Then a comparison has been made with a more realistic model which assumes workers start evacuating according to a prearranged course as soon as hydrogen sulfide detection system alarms. These two models represent the worst and best emergency response effects, respectively, and the analysis results demonstrate significant differences. Results indicate that the CFD and dose–response combined approach is a good way for estimating fatalities of individuals exposed to accidental toxic gas releases.     

A new pre-assessment model for failure-probability-based-planning by neural network

At present, the prediction of failure probability is based on the operation period for laid pipelines, and the method is complicated and time-consuming. If the failure probability can be predicted in the planning stage, the risk assessment system of gas pipeline will be greatly improved. In this paper, the pre-laying assessment model is established to minimize risk of leakage due to piping layout. Firstly, Fault Tree Analysis (FTA) modeling is carried out for urban natural gas pipeline network. According to expert evaluation, 84 failure factors, which can be determined in the planning stage, are selected as the input variables of the training network. Then the FTA model is used to calculate the theoretical failure probability value, and the failure probability prediction model is determined through repeated trial calculation based on BP (Back Propagation Neural Network) and RBF (Radial Basis Function), for obtaining the optimal network parameter combination. Finally, two prediction models are used to calculate the same example. By comparing our pre-assessment model with the theoretical prediction consequences of the fault tree, the results show that the error of RBF prediction model can be close to 3%, which proves the validity and correctness of the method.     

Advanced multi-perspective computer simulation as a tool for reliable consequence analysis

Major accidents involving hazardous materials are a crucial issue for the chemical and process industries. Many accidental events taken place in the past showed that dangerous substances may pose a severe threat for people and property. Aiming at loss prevention, a series of actions have been instituted through international regulations concerning hazardous installations safety preparedness. These actions involve efficient land-use planning, safety studies execution, as well as emergency response planning drawing up. A key factor for the substantial consideration of the above is the effective prediction of possible accident forms and their consequences, for the estimation of which, a number of empirical models have been developed so far. However, (semi-)empirical models present certain deficiencies and obey to certain assumptions, thus leading to results of reduced accuracy. Another approach that could be used for this purpose and it is discussed in this work, is the utilization of advanced computational fluid dynamics (CFD) techniques in certain accident forms modeling. In particular, composite CFD-based models were developed for the simulation of several characteristic accident forms involving isothermal and non-isothermal heavy gas dispersion, confined and unconfined explosion in environment of complex geometry, as well as flammable cloud fire. The simulation cases were referred to real-scale trials allowing us to conclude about the validity of the quantitative results. Comparisons of the computational predictions with the experimental observations showed that obtained results were in good agreement with the experimental ones, whereas the evaluation of statistical performance measures proved the simulations to be statistically valid.     

环己烷泄漏事故模拟与风险控制措施研究

环己烷具有闪点低、爆炸极限宽等特点,一旦发生泄漏,着火爆炸的危险性随时存在。利用计算流体动力学模拟的方法对工程项目中环己烷的泄漏事故进行模拟及风险分析,建立环己酮生产装置的全尺寸三维模型并进行仿真计算,模拟了不同泄漏场景所形成的环己烷可燃气体云团瞬态发展过程及影响范围,并对建构筑物的布局对可燃气体云团的扩散行为的影响进行研究。研究结果表明,通过优化建构筑物布局,可有效降低该装置环己烷的燃爆风险,为企业相关装置的总图布置及环己烷泄漏的安全监控和应急响应提供有价值的参考数据。     

Numerical and experimental study on flame spread over conveyor belts in a large-scale tunnel

Conveyor belt fires in an underground mine pose a serious life threat to the miners. This paper presents numerical and experimental results characterizing a conveyor belt fire in a large-scale tunnel. A computational fluid dynamics (CFD) model was developed to simulate the flame spread over the conveyor belt in a mine entry. Thermogravimetric analysis (TGA) tests were conducted for the conveyor belt and results were used to estimate the kinetic properties for modeling the pyrolysis process of the conveyor belt burning. The CFD model was calibrated using results from the large-scale conveyor belt fire experiments. The comparison between simulation and test results shows that the CFD model is able to capture the major features of the flame spread over the conveyor belt. The predicted maximum heat release rate, and maximum smoke temperature are in good agreement with the large-scale tunnel fire test results. The calibrated CFD model can be used to predict the flame spread over a conveyor belt in a mine entry under different physical conditions and ventilation parameters to aid in the design of improved fire detection and suppression systems, mine rescue, and mine emergency planning.     

尾矿坝浸润线数据挖掘预测模型的样本选取研究

本文分别应用时间序列功能模型和回归模型,在原始数据的基础上建立样本,并运用支持向量回归机算法对样本进行训练,得出了尾矿坝浸润线埋深预测模型并进行了实例应用。研究证明,运用时间序列模型选取训练样本能够得出更为精确的预测结果。     

基于数据填补-机器学习的煤与瓦斯突出预测效果研究*

为解决煤与瓦斯突出事故数据集少,数据缺失严重的问题,提出将多重插补（MI）和随机森林填补(MF)应用于填补缺失参数,并将填补前和填补后的数据输入SVM,ELM,RF 3种机器学习算法进行训练,构建9种耦合模型。采用总体准确率、局部准确率、运行时间这3种指标评价模型性能。研究结果表明:采用数据填补算法后,由于训练样本增大,煤与瓦斯突出事故预测的总体准确率提高,运行时间增长;MF-RF模型的总体准确率与事故预测准确率最高,分别为97.90%和98.93%;RD-ELM模型的运行时间最短,为0.24 s;多重插补使得煤与瓦斯突出预测的总体准确率提高0.98%~1.11%,随机森林填补总体准确率提高5.13%~7.50%,随机森林填补的效果好于多重插补。