人工神经网络在台风预报中的应用初探

为了探索人工神经网络应用于台风分类预报的可能性，用具有Ｓ－特特性函数的Ｂ－Ｐ网络对每类台风的标准样本进行训练，由训练结束后的权值和阈值及待预报样本的因子测值，计算出网络输出值，作出台风分类预报。文中的实例表明人工神经网络用于台风预报具有客观性、可行性和实用性。     

生态足迹动态预测模型构建与分析——以武汉市为例

生态足迹的动态预测能够为区域可持续发展政策的制定提供科学的理论依据,但其研究仍处于探索阶段.以武汉市为例,选取表征社会经济发展的重要指标建立了社会经济系统指标体系,在对武汉市1978～2004年生态足迹时间序列计量的基础上,应用偏最小二乘算法构建了生态足迹动态预测模型,并根据VIP值的大小分析了各指标对生态足迹的影响机制.结果表明,武汉市"十一五"规划期间生态足迹呈上升趋势,其中规划期末2010年的人均生态足迹为2.810 5 hm2,高于全球生态阈值(人均生态足迹为2.200 0 hm2);根据VIP值得出人口、能源消费量和第三产业所占比重为最重要的影响因子,并就如何实现武汉市"十一五"规划目标提出了政策性建议.将科技进步贡献率、能源消费量等重要因子纳入社会经济系统指标体系,弥补了在生态足迹定量测度中忽略社会、经济、科技因素对其影响的不足;另一方面,引入偏最小二乘算法中的VIP值更准确地评价了各指标对生态足迹变化的影响,为生态足迹动态预测研究作了进一步地完善与改进.     

水—能源—粮食纽带关系若干问题解析

水、能源、粮食均为战略性、基础性资源,三者之间相互关联,彼此制约,深刻影响着人类的生产、生活和生态环境。通过梳理水—能源—粮食纽带关系领域的研究内容、研究方法、研究尺度等,进一步明确了变化环境影响下水—能源—粮食纽带关系概念框架及其风险关系的传递与表达,指出当前对水—能源—粮食纽带关系理解尚未统一,广泛采用的有“联系论”和“方法论”。现有研究对水—能源—粮食纽带关系进行了现状评估,资源间相互消耗关系的量化与不同情景下的仿真模拟。研究中存在数据缺失、不一致,评估因素单一,缺少动态反馈机制,难以真正应用到政策调控中等问题。因此,数据整合、模型集成、风险评估与动态调节、提高韧性、城市尺度研究和智慧管理,均是水—能源—粮食纽带关系发展关注的重点领域。     

A new prediction method for dust explosion venting at high static activation pressures

Venting is an effective way to prevent harmful dust explosions, but the existing prediction methods are imprecise and are suitable only for applications with low activation pressures. A new method is proposed for predicting pressures based on an analysis of energy losses at high activation pressures and verified by aluminum dust explosion experiments. Compared with the experimental results, the results of the new model are relatively stable under working conditions with different activation pressures and venting areas. Based on the analysis of energy losses, the changes in the energy loss rate, temperature, and venting velocity during venting are found to be asynchronous. The thermal energy loss, which accounts for over 80 percent of the total, is expected to be larger than the kinetic energy loss. The thermal energy loss rate changes rapidly during venting, while the kinetic energy loss rate remains relatively stable. The new model is more accurate than the NFPA68 standard, which fails to consider the thermal energy loss. Neglecting the thermal energy loss may result in an underestimation of the pressure reduction; this error increases with decreasing activation pressure.     

Degradation of chlordecone and beta-hexachlorocyclohexane by photolysis, (photo-)fenton oxidation and ozonation

Intensive use of chlorinated pesticides from the 1960s to the 1990s has resulted in a diffuse contamination of soils and surface waters in the banana-producing areas of the French West Indies. The purpose of this research was, for the first time, to examine the degradation of two of these persistent pollutants – chlordecone (CLD) and beta-hexachlorocyclohexane (β-HCH) in 1 mg L^?1 synthetic aqueous solutions by means of photolysis, (photo-) Fenton oxidation and ozonation processes. Fenton oxidation is not efficient for CLD and yields less than 15% reduction of β-HCH concentration in 5 h. Conversely, both molecules can be quantitatively converted under UV-Vis irradiation reaching 100% of degradation in 5 h, while combination with hydrogen peroxide and ferrous iron does not show any significant improvement except in high wavelength range (>280 nm). Ozonation exhibits comparable but lower degradation rates than UV processes. Preliminary identification of degradation products indicated that hydrochlordecone was formed during photo-Fenton oxidation of CLD, while for β-HCH the major product peak exhibited C₃H₃Cl₂ as most abundant fragment.     

Effect of inert species on the laminar burning velocity of hydrogen and ethylene

The maximum laminar burning velocity (LBV) of a fuel-air mixture is an important input parameter to vapor cloud explosion (VCE) blast load prediction methods. In particular, the LBV value has a significant impact on the predicted blast loads for high reactivity fuels with the propensity to undergo a deflagration-to-detonation transition (DDT). Published data are available for the maximum LBV of many pure fuel-air mixtures. However, little test data are available for mixtures of fuels, particularly for mixtures of fuels and inert species. Such mixtures are common in the petroleum refining and chemical processing industries. It is therefore of interest to be able to calculate the maximum LBV of a fuel/inert mixture based on the mixture composition and maximum LBV of each component.This paper presents measured test data for the maximum LBV of H₂/inert and C₂H₄/inert mixtures, with both nitrogen and carbon dioxide as the inert species. The LBV values were determined using a constant-volume vessel and the pressure rise method. This paper also provides a comparison of the measured LBV values with simplified LBV prediction methods.     

基于动力学与混合核函数LS-SVM的厌氧发酵产气量预测模型研究

利用农业生产中产生的牲畜粪便牛粪和秸秆进行厌氧发酵试验,以产气量为参考指标,研究牛粪和秸秆厌氧发酵的产气量预测模型。试验过程中测量发酵物的氧化还原电位(ORP)、p H值、挥发性脂肪酸(VFA)、氨氮含量、化学需氧量(COD)、电导率及产气量,并用灰色关联分析法筛选出与产气量关联较强的参数,将筛选出的参数作为多项式核函数和高斯径向基核函数构建的混合核函数LS-SVM模型的输入量,训练出预测能力较强的混合核函数LS-SVM模型,然后在混合核函数LS-SVM模型中融入微生物动力学构建产气量预测模型,并用验证集验证模型性能。仿真结果表明,筛选得到关联较强的参数为ORP、p H值、VFA、氨氮含量;融入动力学的混合核函数LS-SVM模型与混合核函数LS-SVM模型相比,产气量的预测更准确,误差更小,最小误差减少了近一个数量级,试验证明该模型有效。     

铁道线路通行列车数量与铁路边界噪声关系研究

合理的列车编组数量一直是确保铁路边界噪声达标的重要条件。从噪声控制理论出发,对铁路边界噪声达标前提下,线路上所能允许通过的最大列车编组数量(主要是车厢数量)及鸣笛时间进行了有益的探讨,并提出了可用于直接判定铁路边界噪声达标状况的方法,以及计算铁路边界噪声值的简易公式。     

某有色金属企业污水处理厂污水处理工艺

对有色金属冶炼过程中产生的废水按特征因子的不同进行分质处理。其中,高浓度氨氮废水采用三级氨氮吹脱、吸收工艺进行脱氨预处理;含砷酸性废水采用三段中和-铁盐混凝法预处理;经预处理废水混合后,采用石灰法分级沉淀处理废水中重金属离子;然后根据各单位对回用水水质的不同要求,对废水进行深度处理,实现了污水处理后全部回用,污水零排放的目的。并且对污水中的氨、石膏、镍等资源进行回收利用,为企业降低了运行成本,并且防止了二次污染。     

基于MaxEnt的加拿大一枝黄花在中国的潜在分布区预测

基于采样后的加拿大一枝黄花(Solidago canadensis)在北半球的分布数据,以温度和降水作为预测的环境因子,以MaxEnt模型作为物种适生性预测模型,预测加拿大一枝黄花在我国的潜在分布。结果表明:加拿大一枝黄花在我国有广泛的潜在分布区,目前的实际分布远未达其最大潜在分布范围,仍有可能继续扩散。加拿大一枝黄花主要适生区包括上海、江苏、浙江、安徽、山东、河南、湖北、湖南、江西、福建、重庆、贵州、四川省东部、广东和广西2省北部、云南省东北部、山西和陕西2省南部地区。必须采取有效措施阻止加拿大一枝黄花扩散蔓延。