基于影响因素分析的给水管网主体水余氯衰减模型

余氯在管网中主要发生主体水反应和管壁反应。一方面自身会不断消耗衰减,导致管网末梢处余氯值达不到要求,另一方面,氯会与水体中的某些有机物反应生成三卤甲烷等消毒副产物。以上作为氯消毒的两个主要问题,既相互矛盾又相互联系。其中余氯在主体水中的反应受到多种因素的影响,这些因素包括温度,初始余氯浓度和水体水质情况等等。论文在正交试验设计的基础上,通过对试验结果的极差和方差分析,得出结论:温度是影响余氯衰减的最主要因素,其次是水体水质情况,影响最小的是初始余氯浓度。在正交试验结果的基础上做定性分析的同时,采用编程拟合建立了改进的主体水余氯衰减模型,得到的参数组合为A=21 428 571,m=0.539,α=2.903。该改进模型不仅能很好地拟合试验结果,而且由于是基于正交试验所建立,模型的代表性和适用性都更加优越。     

采用功率谱方法提取地球自由振荡信息

利用中国数字地震台网的三个台站所记录到的2011年日本里氏9.0级地震的地震数据,在没有进行去固体潮的前提下进行功率谱分析,获得了地球球型自由振荡振型,通过与PREM模型相对比,发现所提取的94个振型,包括76个基频振型（0S0,2S0～0S76）和18个谐频振型,与理论值的误差很小,所有数据的平均偏差为0.14%。文中发现该地震激发的0S2,0S3,0S4和1S2自由振荡振型有明显的频谱分裂现象。     

活性碳纤维-过硫酸盐体系处理焦化废水生化出水的实验研究

采用活性碳纤维（ACF）活化过硫酸盐（PMS）深度处理焦化废水生化出水,并考察了PMS浓度、ACF投加量、pH值对焦化废水生化出水中COD和色度去除效果的影响.结果表明,ACF/PMS体系比ACF/PS体系更能有效去除焦化废水中的有机物和色度.在PMS浓度为8.0 mmol·L^-1、ACF投加量为4.0 g·L^-1、pH约为8.0、温度为30 ℃的条件下,反应120 min后焦化废水生化出水中COD和色度的去除率分别达到88.7%和91.2%.ACF表征分析和重复利用实验结果说明,ACF具有较好的吸附和催化性能,且重复使用稳定性较好.自由基鉴定实验表明,SO₄^-· 和·OH均为反应体系中起主导作用的活性自由基.三维荧光光谱显示,焦化废水生化出水中的类富里酸和类腐殖酸物质可被有效降解转化为小分子物质.     

紫外耦合游离亚硝酸强化剩余污泥厌氧发酵产酸研究

为打破传统厌氧发酵过程中污泥破壁溶胞困难和产酸效能低的瓶颈,探究了紫外光（UV）耦合游离亚硝酸（FNA）预处理对污泥发酵产酸的影响,并与热（H）和超声法（US）耦合FNA预处理进行了对比分析.结果表明,UV辅助FNA联合预处理（FNA-UV）对细胞破碎和胞外聚合物的剥离具有协同效应,·OH和·O₂^-作为反应中间体,其强度远高于其他预处理组（FNA-US和FNA-H）,与·NO,·NO₂及ONOO^-等中间产物共同促进了溶解性有机物的释放,溶解性碳水化合物和蛋白质含量相比FNA组分别提升60%和90%,进而为后续水解产酸过程提供了充足的底物.FNA-UV组短链脂肪酸（SCFAs）浓度于第4d达到峰值,为（201.8±4.8） mg COD/g VSS,相比FNA组提升67%,乙酸占比高达56.8%.通过发酵末期对各体系进行碳平衡分析表明,紫外耦合FNA预处理在污泥减量、溶解性有机物的释放与转化、SCFAs的产生方面具有重要作用.微生物群落分析表明,FNA-UV对功能菌群的富集发挥重要作用,表现为厌氧发酵菌和反硝化菌的有效增强,相比其他各组提升了23.7%~270.6%.     

Predators choosing between patches with standing crop: the influence of switching rules and input types

Where prey arriving in a patch are not consumed immediately, they will accumulate. Predators are then presented with a prey density or standing crop that increases through further input, and decreases through the consumption by predators. Firstly, I show that the switching rule of predators has a significant influence on the expected predator equilibrium distribution in such a dynamic system. Three rules are compared; for all rules, analytical solutions are calculated (where possible). To test their plausibility for natural situations, predator distributions are simulated given the assumption that each predator obtains individual patch profitability estimates by using a common learning rule. As long as prey arrive in the patches in constant numbers per time unit, the first rule leads to input matching because predators stop switching when consumption in the two patches is equal. The other two rules, where predators continue to sample both patches even in the equilibrium state, lead to predator distributions where the more profitable patch is underused. The final equilibrium depends on the exact assumptions of the switching rule; however, it is independent of interference. But if the input delivered into a patch is a function of the current prey standing crop (for example in a reproducing prey population), predator and prey distributions will not reach an equilibrium in most cases: either standing crops increase indefinitely, or they approach zero, with all predators concentrating on the better patch. Only a small number of parameter sets show intermediate crops that are reasonably stable. With this input type, only up to 54% of the simulations reach the expected distribution. In a system with competition for dynamic standing crop, it is therefore essential to know the type of input and the switching-rule used by predators to be able to predict equilibrium predator distributions. Received: 17 March 1995/Accepted after revision: 5 November 1995