BASIN SCALE WATER MANAGEMENT AND FORECASTING USING ARTIFICIAL NEURAL NETWORKS1

ABSTRACT: Water scarcity in the Sevier River Basin in south‐central Utah has led water managers to seek advanced techniques for identifying optimal forecasting and management measures. To more efficiently use the limited quantity of water in the basin, better methods for control and forecasting are imperative. Basin scale management requires advanced forecasts of the availability of water. Information about long term water availability is important for decision making in terms of how much land to plant and what crops to grow; advanced daily predictions of streamflows and hydraulic characteristics of irrigation canals are of importance for managing water delivery and reservoir releases; and hourly forecasts of flows in tributary streams to account for diurnal fluctuations are vital to more precisely meet the day‐to‐day expectations of downstream farmers. A priori streamflow information and exogenous climate data have been used to predict future streamflows and required reservoir releases at different timescales. Data on snow water equivalent, sea surface temperatures, temperature, total solar radiation, and precipitation are fused by applying artificial neural networks to enhance long term and real time basin scale water management information. This approach has not previously been used in water resources management at the basin‐scale and could be valuable to water users in semi‐arid areas to more efficiently utilize and manage scarce water resources.     

西昌高原鳅人工繁殖技术研究

对西昌高原鳅(Triplophysa xichangensis)人工繁殖中的催产药物及其剂量、受精方式、孵化方法和仔鱼水花培育方法等一系列关键技术进行了详细研究。结果表明,对性腺发育至Ⅳ+~Ⅳ++期的雌鱼分二次注射LRH-A2 0.1μg/g(鱼体重)和HCG 20 IU/g(鱼体重),雄鱼剂量减半,注射量0.1 mL,能很好地促进雌雄鱼发情并产卵排精,在水温19±0.5℃时的效应时间为30.5~31 h;在一个特制的塑料盆中完成了亲鱼培育、打针催产、发情、自然产卵受精、孵化以及仔鱼培育等全部人工繁殖过程,受精率、孵化率和仔鱼水花培育20 d的成活率平均分别为97.2%、85%和85%。本研究成果可应用于西昌高原鳅规模化繁育,对长江上游鱼类生物多样性保护有着积极意义。     

Oxygen demand for the stabilization of the organic fraction of municipal solid waste in passively aerated bioreactors

Conventional aerobic waste treatment technologies require the use of aeration devices that actively transport air through the stabilized waste mass, which greatly increases operating costs. In addition, improperly operated active aeration systems, may have the adverse effect of cooling the stabilized biomass. Because active aeration can be a limiting factor for the stabilization process, passive aeration can be equally effective and less expensive. Unfortunately, there are few reports documenting the use of passive aeration systems in municipal waste stabilization. There have been doubts raised as to whether a passive aeration system provides enough oxygen to the organic matter mineralization processes. In this paper, the effectiveness of aeration during aerobic stabilization of four different organic fractions of municipal waste in a reactor with an integrated passive ventilation system and leachate recirculation was analyzed. For the study, four fractions separated by a rotary screen were chosen. Despite the high temperatures in the reactor, the air flow rate was below 0.016 m³/h. Using Darcy’s equation, theoretical values of the air flow rate were estimated, depending on the intensity of microbial metabolism and the amount of oxygen required for the oxidation of organic compounds. Calculations showed that the volume of supplied air exceeded the microorganisms demand for oxidation and endogenous activity by 1.7–2.88-fold.     

基于人工神经网络的气体泄爆最大超压预测研究

为解决传统经验公式在预测气体泄爆中最大超压出现时的较大偏差或过于保守的问题，提出使用人工神经网络预测气体泄爆最大超压。基于124组实验数据，采用BP与RBF神经网络，通过优化算法计算与迭代循环对泄爆样本中的影响因素进行降维与选择，并确定2类神经网络本身在学习与计算气体泄爆样本时的相关参数。结果表明:PCA（主成分分析法）在当前样本条件下的降维效果较差，而通过迭代对比确认气体泄爆样本中的5类特征全部保留时神经网络的训练模拟效果最好；通过对124组实验数据进行随机挑选训练集与测试集的训练模拟结果发现，神经网络对气体泄爆中最大超压的预测效果较好；通过对比Molkov提出的和经Fakandu等改进的NFPA 68经验公式以及2类神经网络的预测结果表明，神经网络相比于传统气体泄爆经验公式具有明显优势。     

Efficacy of created and restored nesting sites for the conservation of colonial Laridae in the South of France

By rapidly modifying key habitat components, habitat restoration is at risk of producing attractive cues for animals without providing habitats of sufficient quality. As such, individual fitness components, such as reproduction, could be reduced and restored habitats could become ecological traps. This risk notably appears by using artificial constructions in restoration projects, yet few studies have evaluated their efficacy in a robust way. We investigated this by analyzing 154 islets that were created or restored to improve the conservation status of 7 colonial Laridae species in the South of France. From 2007 to 2016, we compared occupancy dynamics and breeding parameters of these species between the restored sites and 846 unmanaged nesting sites. We also explored species’ preference for different nesting site characteristics and their respective effect on breeding parameters. Restored nesting sites were 2–9 times as attractive as unmanaged sites for all species except the Black-headed Gull (Chroicocephalus ridibundus). Colonization probability was up to 100 times higher in sites already used by other species the previous year and increased with distance to the shore until >0.2 when distance was over 250 m. Abandonment probability was 29–70% lower when breeding was successful the previous year in all species except the Sandwich Tern (Thalasseus sandvicensis). Productivity and breeding success probability were 2 times higher on managed sites. Distance from the shore was an important attractive characteristic of artificial nesting sites in all species. Other nesting site characteristics had species-specific effects on colonization, abandonment, and breeding success. Our results indicate that managed nesting sites are successful conservation tools for colonial Laridae in the Mediterranean and do not act as ecological traps. Our study showed that testing the ecological trap hypothesis is a robust way to evaluate the success of restoration projects of breeding habitats.     

AAO工艺低氧条件下的运行及其模拟

利用厌氧-缺氧-好氧(AAO)工艺探讨了在好氧池低溶解氧(DO)浓度条件下的水质指标变化情况.结果表明,当好氧池DO浓度从2.00 mg·L~(-1)降低到1.00 mg·L~(-1)和0.50 mg·L~(-1)时,系统仍然具有良好的除磷脱氮效果,出水水质指标均满足我国《城镇污水处理厂污染物排放标准》(GB 18918-2002)中的一级A标准.在此基础上,基于Bio Win 4.1软件建立了AAO工艺的活性污泥模型,对模型参数进行了灵敏度分析,并利用动态模拟对模型参数中的聚磷菌吸收乙酸或丙酸的聚羟基烷酸(PHA)产率系数(YP/PHA,seq)、聚磷菌好氧氧化PHA的储磷率(YP/PHA,aerobic)、氨氧化菌的最大单位生长速率(μmax,A)和亚硝酸盐氧化菌的最大单位生长速率(μmax,N)进行了校验.此外,对系统的曝气能耗进行了模拟评估,结果表明,与好氧池DO浓度为2.00 mg·L~(-1)时相比,好氧池DO浓度为1.00 mg·L~(-1)和0.50 mg·L~(-1)时的空气流量可分别节省23.8%和38.1%,氧转移效率可分别提高7.2%和11.7%.     

罕见冬季降雨中人工增雨微物理过程分析研究

用Micaps实时资料、CPAS卫星反演分析资料、NCEP再分析资料、ECMWF再分析资料,多普勒雷达PUP产品资料,对2015年12月1日夜间至2日凌晨沈阳地区罕见冬季降雨过程,进行了天气学分析、及大气微物理人工增雨作业条件分析,并结合沈阳地区人工增雨作业指标,得出本次降雨过程在厄尔尼诺正在发生的背景下,为常见的高空槽转东北冷涡型,水汽、动力等微物理作业条件也适合于开展人工增雨作业,并给出了相应的作业指导参数,为沈阳地区今后在冬季降雨天气过程中开展人工增雨作业积累了宝贵经验.     

ABC算法优化SVR的磨损故障预测模型

目的为了提高故障预测的精度,针对支持向量回归SVR(Support vector machine for regression,SVR)参数选择困难的问题,提出一种采用人工蜂群(artificial bee colony,ABC)算法优化支持向量回归(SVR)的故障预测模型(ABC-SVR)。方法该模型先对样本数据进行重构,然后将故障预测误差(适应度)作为优化目标,通过ABC算法寻优找到最优的SVR参数,建立故障预测模型。最后通过实例仿真验证模型的优越性。结果采用ABC算法优化的SVR故障预测模型进行时间序列预测,能够较好地跟踪发动机滑油金属元素浓度的变化过程,并且能够提前2个取样时间预测异常情况的出现。结论 ABC-SVR模型有效解决了SVR参数选择难题,能够更加准确地表现故障变化规律,提高了故障预测精度。     

两段式曝气对好氧颗粒污泥脱氮性能的影响

基于4个不同进水条件的小试实验探究,得出所接种好氧颗粒污泥表面异养菌在曝气初期阶段大量消耗碳源,对于NO-2-N和NO_3~--N的反硝化效率较低.为此,实验在常温(20~23℃)条件下,在同一周期内先采用低曝气量曝气之后再用高曝气量曝气的两段式曝气方式运行好氧颗粒污泥SBR反应器,低曝气量时长分别取1、2、3 h这3个阶段分别运行,并运用扫描电镜(SEM)和荧光原位杂交技术(FISH)对颗粒污泥进行分析,结果表明AGS粒径增大,反硝化能力提升,NO-2-N的反硝化速率(以LVSS计)在低曝气时长为2 h时升至最高,达9.66 mg·(g·h)~(-1).亚硝态氮积累率不断升高至77.84%,总氮去除率最高达70%.颗粒污泥内部孔隙增多,且细菌多以球菌、椭球状及杆菌为主,氨氧化菌(AOB)占总菌的比例由13.70%升至15.40%.因此,通过两段式曝气过程实现了短程同步硝化反硝化过程并具有较好的脱氮性能.     

ASM1耦合曝气模型对污水处理厂的模拟研究

为了更好地模拟污水处理厂实际的曝气方式及溶解氧浓度,采用更加精确的曝气模型对某污水处理厂进行了模拟研究.该模型考虑了污泥浓度、堵塞及温度等因素对氧传质的影响,并将该曝气模型与ASM1相耦合.首先,通过灵敏度分析确定了对该污水处理厂出水指标影响较大的参数,分别为异养菌产率系数YH、异养菌最大比增长速率μH、自养菌最大比增长速率μA及自养菌氧利用半饱和系数KOA.然后,测定部分灵敏度较高的参数并使用测量结果对模型进行校正,校正结果使用希尔系数表征.最后,使用另一批数据对校正参数进行验证.模拟的COD、氨氮、总氮和好氧池末端溶解氧的希尔不等系数分别是0.13、0.23、0.20和0.15,而通过控制溶解氧浓度或理论氧转移效率模型模拟则出现氨氮、溶解氧、总氮及COD的希尔不等系数不同程度上升的结果.校验后的ASM1/曝气耦合模型参数可以很好地应用于污水处理厂的模拟当中,对于优化污水厂曝气控制具有重要意义.