边坡稳定性自动机器学习预测方法研究

为了简便有效地评估边坡稳定性状态,针对目前传统机器学习的算法选择与超参数优化等难题,提出1种边坡稳定性自动机器学习预测方法。首先,简要介绍5种主流开源自动机器学习框架;其次,以422组边坡稳定性样本为数据集,进行自动机器学习纯自动化训练,并与传统机器学习对比分析模型的性能与耗时;最后,综合讨论与比较典型自动机器学习框架的特性。研究结果表明：自动机器学习预测效果总体上优于传统机器学习,提升边坡稳定性预测准确率和稳健性,且无需人为干预。研究结果可为岩土工作人员准确可靠地评价边坡稳定性提供便捷条件。     

压缩空气泡沫水平管路压力损失数值模拟研究

为厘清压缩空气泡沫水平管道输运压力衰减规律,考虑压缩空气泡沫实际工程运用,利用STAR-CCM软件研究泡沫液种类、混合比和管径对湿泡沫水平管网输运过程中压力衰减的影响规律,并结合理论分析建立压缩空气泡沫水平输运过程中的压力衰减预测模型。研究结果表明：管路压降与泡沫液黏度呈正相关性,在混合液流量270 L/min,气体流量1 750 L/min条件下,相同管径高黏度抗溶性水成膜泡沫(AFFF/AR)压降约为低黏度A类和水成膜泡沫(AFFF)1.3倍;对于相同类型泡沫,混合比对管路压降影响较小,100 m长90 mm管径不同混合比之间最大压降差值约为7.21 kPa;管径对压降影响较大,相同泡沫条件下,50 mm管径压降是80 mm管径压降的约9.4倍,80 mm管径压降是100 mm管径压降的约2.8倍,当管径大于80 mm时,不同泡沫对压降的影响逐步减小。压力衰减预测模型计算的压降值与模拟值较前人开展研究所得实验值误差在18%以内,研究结果可以为压缩空气泡沫水平管网设计提供一定理论参考。     

Toxicity Testing of Leachate from Waste Landfills Using Medaka (Oryzias Latipes) for Monitoring Environmental Safety

To investigate the environmental safety of waste disposal landfill sites and of land reclaimed from such sites, we evaluated the toxicity of leachate from these sites by a combination of bioassays in the Japanese killifish medaka Oryzias latipes. We tested for lethal toxicity in adult and larval medaka and for hatching inhibition of embryos from eggs. As biochemical evidence of the effects of leachate exposure, CYP1A (EROD activity) and vitellogenin (Vtg) were induced. We also bioassayed water-treated leachate and downstream river water. Leachate solution was lethal to larval and adult medaka. Embryo hatchability was inhibited, and abnormal hatching, spinal deformity and anisophthalmia occurred in embryos exposed to leachate solution. CYP1A was induced by exposure to leachate solution diluted to 1.0%, and EROD activity was significantly higher than in control. Vtg and unknown proteins were induced in the sera of male medaka exposed to the diluted leachate solution. Conventional water treatments worked effectively to remove toxic compounds but did not work well to remove element ions, including heavy metals. Treated leachate produced neither lethal toxicity nor hatching abnormalities during the exposure period. Fish toxicity tests for leachate would be useful for monitoring the environmental safety of landfill sites.     

Relationships Between Toxicity and Concentrations of Chemical Contaminants in Sediments from Sydney Harbour, Australia, and Vicinity

Correlation analyses between measures of toxicity and concentrations of chemical contaminants were conducted for 103 surficial sediments from Sydney Harbour, Australia, and vicinity. Toxicity tests consisted of amphipod survival and reburial tests of whole sediments (Corophium colo), sea urchin fertilisation and larval development tests of pore waters (Heliocidaris tuberculata) and microbial bioluminescence (Microtox) tests of solvent extracts and pore waters. Toxicity in most tests correlated with concentrations of metallic contaminants, in particular, zinc, lead and copper. Organic contaminants did not correlate as significantly with toxicity. However, Heliocidaris tuberculata showed relationships with organochlorine compounds in samples with low to moderate metals contamination. Toxicity in the Microtox solvent extract test appeared to be primarily influenced by the presence of sulfur. This study has no precedent in Australia and the results support the validity of using local indigenous species in toxicity tests of field-collected sediments. This toxicity/chemistry dataset may be used in evaluations of sediment quality guidelines recently introduced to Australia.     

Evaluation of Traffic Noise Pollution in Amman, Jordan

The City of Amman, Jordan, has been subjected to persistent increase in road traffic due to overall increase in prosperity, fast development and expansion of economy, travel and tourism. This study investigates traffic noise pollution in Amman. Road traffic noise index L ₁₀(1 h) was measured at 28 locations that cover most of the City of Amman. Noise measurements were carried out at these 28 locations two times a day for a period of one hour during the early morning and early evening rush hours, in the presence and absence of a barrier. The Calculation of Road Traffic Noise (CRTN) prediction model was employed to predict noise levels at the locations chosen for the study. Data required for the model include traffic volume, speed, percentage of heavy vehicles, road surface, gradient, obstructions, distance, noise path, intervening ground, effect of shielding, and angle of view. The results of the investigation showed that the minimum and the maximum noise levels are 46 dB(A) and 81 dB(A) during day-time and 58 dB(A) and 71 dB(A) during night-time. The measured noise level exceeded the 62 dB(A) acceptable limit at most of the locations. The CTRN prediction model was successful in predicting noise levels at most of the locations chosen for this investigation, with more accurate predictions for night-time measurements.     

钻井废液毒性检测与评价研究

以几种典型钻井液体系为研究对象,分别选用成熟可靠的检测方法对其生物毒性、化学毒性和重金属含量进行分析,结果表明:部分钻井废液具有一定生物和化学毒性,其中,磺化体系钻井液在生物毒性和化学毒性方面均超标,应重点加强管理;低密度无固相体系和磺化体系钻井液中Hg和Cr元素存在不同程度的超标。     

循环冷却水浓缩倍率对凝汽器管路的腐蚀和结垢趋势

利用水质判断指数来预测不同循环冷却水浓缩倍率下水的结垢趋势,确定形成最佳碳酸钙保护膜所对应的浓缩倍率,以达到最经济的防止循环冷却水系统腐蚀和结垢的目的.     

Tissue accumulation of aluminium is not a predictor of toxicity in the freshwater snail, Lymnaea stagnalis

The amount of toxic metal accumulated by an organism is often taken as an indicator of potential toxicity. We investigated this relationship in the freshwater snail, Lymnaea stagnalis, exposed to 500 μg l⁻¹ Al over 30 days, either alone or in the presence of phosphate (500 μg l⁻¹ P) or a fulvic acid surrogate (FAS; 10 mg l⁻¹ C). Behavioural activity was assessed and tissue accumulation of Al quantified. Lability of Al within the water column was a good predictor of toxicity. FAS increased both Al lability and behavioural dysfunction, whereas phosphate reduced Al lability, and completely abolished Al-induced behavioural toxicity. Tissue accumulation of Al was not linked to toxicity. Higher levels of Al were accumulated in snails exposed to Al + P, compared to those exposed to Al alone, whereas FAS reduced Al accumulation. These findings demonstrate that the degree of tissue accumulation of a metal can be independent of toxicity.     

信息数字化与雾的细化预测

在先前工作的基础上,尝试了信息数字化细化取样的细化分析方法.结果表明:雾的分析和预测效果更能得到"落点"预测的实效性改善.     

Effects of low concentrations of the phenylurea herbicide diuron on biofilm algae and bacteria

A system of recirculating channels was used in this study to examine the long-term effects (29 d) of environmentally realistic concentrations of the herbicide diuron (from 0.07 to 7 μg L⁻¹) on biofilm communities. The autotrophic activity of biofilms was affected by this herbicide, as reflected by a marked decrease in the photosynthetic efficiency. Diuron exposure also increased chlorophyll-a content and reduced the biovolume of diatom taxa at low concentrations. The effects on bacteria were also remarkable. Bacterial abundance was reduced after a week of exposure to the herbicide at a range of concentrations. Effects were on the number of live bacteria and on the increase in the leucine-aminopeptidase activity. It is suggested that inputs of herbicides to the river ecosystem at low concentrations may cause a chain of effects in the biofilm, which include inhibitory effects on algae but also indirect effects on the relationships between biofilm components.