Effects of combinations of malathion and cypermethrin on survivability and time of metamorphosis of tadpoles of Indian cricket frog (Fejervarya limnocharis)

This study was undertaken to determine the effect of environmentally realistic concentrations of two commonly used pesticides viz., malathion and cypermethrin, using a fully 3 × 3 factorial experiments on the survivability and time of metamorphosis in a common rice paddy field frog (cricket frog) Fejervarya limnocharis under laboratory conditions. The results suggest that cypermethrin is more toxic than malathion and combinations of higher concentrations of cypermethrin (50 μg/L) with malathion (250 and 500 μg/L) are more deleterious to the survivability of tadpoles. With increasing cypermethrin concentration, the survivability of tadpole decreased (r = ?0.986, P = 0.108). But cypermethrin alone induced early metamorphosis among the surviving tadpoles. However, there was a delay in the time required for metamorphosis induced by malathion and its combination with cypermethrin. The delay in metamorphosis may indicate the altered physiological fitness of the individual. The emergent froglets will be subjected to environmental stressors like high temperature and less humidity of post-monsoon tropical climate that could enhance negative influence triggered by pesticides.     

灾难事件后继风险决策

对小概率/高风险的灾难事件的后继风险决策进行探讨,着重在决策的认知神经经济模型框架下,讨论灾难事件后继风险决策的影响因素。灾难事件后人们的风险觉知、信念和假定、经验等认知因素以及灾后情绪因素对于决策分别有不同的影响;人们灾后决策的特点为决策更多直觉化而较少运用理性分析;灾难后继决策有随时间变化的趋势,灾难事件后人们为寻求安全感而出现"损失偏差"(loss bias),该偏差在"获得"和"损失"两种条件下的影响是不同的。灾后风险决策的内在机制或可用思维和决策的双加工系统(dual-process)模型来解释。最后展望了灾难事件后继决策进一步的研究方向。     

基于缓冲区重采样的LSTM林火预测模型*

为提高林火风险预测精度,挖掘地图上隐含的空间信息、时间序列上隐含的长期趋势和循环波动,提出1种基于缓冲区重采样的长短期记忆（LSTM）林火预测模型,选取15个与林火相关的影响因素,以方差膨胀因子为评价指标对其进行多重共线性检验,方差膨胀因子大于10的因素具有共线性,并采用信息增益率验证筛选结果的合理性。考虑到火灾的空间聚集特性,采用缓冲区分析与过采样相结合方法减少样本不均衡现象的影响,最终得到176 732条样本。对12个影响因素和研究时间段的火点建立LSTM预测模型,对森林火灾发生风险进行预测。研究结果表明:基于缓冲区重采样的LSTM林火预测模型有效考虑时空上隐含的信息,预测模型准确率为87.06%,特异性为97.99%,敏感度为76.12%,阳性预测率为97.43%,阴性预测率为80.41%,ROC曲线与AUC值均优于随机森林（RF）和支持向量机（SVM）这2种基准算法。维尔克松秩和检验发现,本文提出的模型与基准算法结果具有显著性差异。研究结果可为提高林火风险预测精度提供参考。     

Improved models of failure time for atmospheric tanks under the coupling effect of multiple pool fires

Fire accidents of chemical installations may cause domino effects in atmospheric tank farms, where a large amount of hazardous substances are stored or processed. Pool fire is a major form of fire accidents, and the thermal radiation from pool fire is the primary hazard of domino accidents. The coupling of multiple pool fires is a realistic and important accident phenomenon that enhances the propagation of domino accidents. However, previous research has mostly focused on the escalation of domino accidents induced by a single pool fire. To overcome the drawback, in this study, the failure of a storage tank under the coupling effect of multiple pool fires was studied in view of spatial and temporal synergistic process. The historical accident statistics indicated that the accident scenario of two-pool fires accounted for 30.6% in pool fires. The domino accident scenario involving three tanks is analyzed, and the typical layout of tanks is isosceles right triangle based on Chinese standard “GB50341-2014”. The thermal response and damage of a target tank heated by pool fires were numerically investigated. The volume of 500 m³, 3000 m³, 5000 m³ and 10000 m³ were selected. Flame temperature was obtained by FDS, and then was input onto the finite element model. The temperature field and stress field of target tanks were simulated by ABAQUS. The results showed that the temperature rise rate of the target tanks under multiple pool fires was higher than that under a single pool fire. The failure time of the tank under the coupling effect of multiple fires was lower than that under the superposition of multiple fires without the first stage. The stress and yield strength were compared to judge the failure of the target tank. The model of failure time for the tank under the coupling effect of pool fires was established. Through the verification, the deviation of this model is 4.02%, which is better than the deviation of 15.76% with Cozzani's model.     

基于分形学的瓦斯爆炸事故时序数据分析模型及应用

针对煤矿瓦斯爆炸灾害子系统的复杂非线性动力学特性,以南方某省1958—2007年的煤矿瓦斯爆炸事故发生数构建时间序列,应用霍斯特时间序列数据的标度行为,提出瓦斯爆炸时间序列的重标极差分析(R/S分析)方法。结果表明:瓦斯爆炸时间序列的霍斯特指数为0.550 2,分形维数为1.449 8,其分式布朗运动轨迹表现出一定的持久性和非高斯性。为诊断瓦斯爆炸事故时间序列变异年份,对时间序列进行分段R/S分析,获得变异年份为1974年,以1974年为分界点构成2个时间序列,霍斯特指数分别为0.589 1和0.697 5,分形维数为1.410 9和1.302 5,与煤矿瓦斯爆炸事故时间序列的持久性趋势一致。煤矿瓦斯爆炸事故时间序列看似无序却蕴藏规律,反映了系统的动力学变化特征,R/S分析方法可以描述煤矿瓦斯爆炸事故时间序列的动力学特征。     

不同墩高下连续刚构桥抗震性能分析

以某大跨连续高墩刚构桥为依托,采用Midas/civil软件建立仿真模型,通过时程分析方法研究双肢薄壁墩参数在地震响应下对高墩连续刚构桥抗震性能的影响.以桥墩高度为主要参数,以模型阵型频率、桥墩墩顶、墩底位移及内力变化三个方面对其抗震性能进行评价.研究结果表明,随着其中一个桥墩高度的增加,桥梁频率不断减小,矮墩墩底及矮...     

火灾中闭式喷头响应时间的概率分布特性研究

喷淋系统中闭式喷头响应时间受火灾增长因子、径向安装距离和喷头响应时间指数等因素的影响。通过对喷头响应时间的模拟数据的统计分析,建立了包含火灾增长因子、径向距离和喷头响应时间指数等参数的喷头响应时间模型。考虑火灾增长因子和径向距离的随机性,获得了喷头响应时间的概率分布特性。研究结果表明,径向距离和火灾增长因子对喷头响应时间的分布特性影响显著。随着火灾增长因子的增大和径向距离的缩小,喷头在较短时间内响应的概率增大。随着径向距离的缩小,喷头响应时间指数对喷头响应时间分布的影响变小。研究可为优化喷头布置和评估喷头响应的有效性提供分析工具。     

基于人的认知可靠性模型(HCR)的接车进路办理诊断时间的研究

接发列车作业事故在铁路运输事故中占有相当大的比重。为减少接车作业事故的发生,合理制定车站接车进路办理时的诊断时间是保证接车作业安全的关键。针对车站采用写实法查定时具有较大随意性的问题,提出采用人的认知可靠性(HCR)模型,根据人因失误概率计算接车作业诊断时间的方法,并在自动闭塞区段通过模拟机实验进行了验证。结果表明,HCR模型确定的诊断时间与人因失误之间的关系与写实法的测定数据具有良好的符合性;作业人员的训练水平对诊断时间有较大的影响;利用该模型,根据作业人员的训练水平与可接受的人因失误概率,可以确定接车进路办理的最佳诊断时间,或对写实法测定的诊断时间进行订正。     

Influence of super-absorbent polymer on the growth rate of gas hydrate

The growth rate of hydrate and morphology of methane hydrate formation were studied in a visual pressure cell at 5.5 MPa. The gas hydrate formation was carried out (coal mine methane (CMM) + tetrahydrofuran (THF) + sodium dodecyl sulphate (SDS) + H₂O) with and without SAP. Experimental data on the hydrate growth rate and induction time were obtained for three different CMM samples. The influence of SAP on hydrate growth rate was determined. Results showed that after the addition of SAP, with the methane concentration increased in CMM, the induction time was reduced by 9 min, 10 min and 3 min, and the growth rate was shortened by 0.56 × 10⁻⁶/m³ min⁻¹, 0.53 × 10⁻⁶/m³ min⁻¹ and 1.42 × 10⁻⁶/m³ min⁻¹, respectively. This study could be useful for the recovery of methane from CMM by forming hydrate in the chemical and mining industry.     

模式搜索算法在毒气泄漏中的源强反算

基于泄漏源下风向的浓度监测数据并结合大气扩散模式建立反算模型,以确定泄漏源的位置和强度。以扩散模式仿真的浓度数据与监测数据的匹配度作为目标函数,将反演问题转化为优化问题,利用模式搜索算法迭代优化。以高斯模型为例验证了算法的可行性,结果表明利用探测器提供的测量浓度值,模式搜索算法能够在较短时间内搜索到最优解,在计算复杂性或时间上较梯度型算法和智能优化算法有一定优势。该算法能够及时而准确地反算出泄漏源强度和位置,为事故的应急响应与救援提供依据。