太湖流域上游平原河网污染物综合衰减系数的测定

改善太湖水质需要削减上游河流进入太湖的污染物总量.为了探求太湖流域上游平原河网的自净能力,开展原位实验测定了枯水期高锰酸盐指数、氨氮(NH_4~+-N)、总氮(TN)和总磷(TP)的综合衰减系数,根据河道的水力特征对综合衰减系数进行了修正,并利用一维稳态水质模型对修正前后综合衰减系数的可靠性进行了验证.结果表明,高锰酸盐指数、NH_4~+-N、TN和TP的综合衰减系数分别为:0.0296~0.4106、0.0224~0.3564、0.0137~0.3046和0.0555~0.5725 d~(-1).可靠性验证表明高锰酸盐指数、NH_4~+-N、TN和TP综合衰减系数修正前的平均相对误差分别为8.39%、14.40%、11.43%和19.22%,修正后的平均相对误差分别为10.65%、14.34%、11.37%和19.24%.修正前后高锰酸盐指数、NH_4~+-N、TN和TP的平均相对误差均小于20%且变化不显著,表明综合衰减系数的测定结果能够为太湖流域上游平原河网的污染物总量控制管理提供科学参数;也表明枯水期的水力条件对综合衰减系数的影响较小.     

路面径流水环境下交通承载力分析

为了对路面径流水容许污染总量控制下的交通承载力问题进行探讨,利用神经网络具有的非线性映射能力和遗传算法具有的全局随机搜索能力,结合公路路面径流水质检测数据,提出了一种基于遗传神经网络进行公路交通环境承载力反计算的分析方法,应用该方法可根据路面径流水质污染数据反演出路段交通量大小,并可据此进行交通量与路而径流水质污染的关...     

人工智能在金属塑性加工领域中的应用

概述了人工智能及其 2个主要分支 (专家系统和人工神经网络 ) ,讨论了遗传算法与专家系统集成以及神经网络与模糊逻辑、专家系统集成的必要性和集成方法 ,分别介绍了专家系统和人工神经网络在塑性加工领域中的应用现状。     

基于GA-ELM浆体管道输送临界流速预测模型研究

针对浆体管道输送临界流速预测难度大、精确度低等技术难题,提出了基于极限学习机(ELM)的临界流速预测模型,用训练集对模型进行训练,以验证集预测值的均方误差作为适应度函数,利用遗传算法(GA)对ELM模型参数进行优化,应用优化得到的ELM模型对预测集进行预测。以某矿山为例,模型参数优化结果如下:隐含层节点数L为400,输入权值ai、偏置向量bi最优组合下预测结果适应度为0.0201。采用优化的ELM模型对预测集进行预测,预测结果的最大相对误差x=3.96%,平均相对误差y=1.58%,对比BP神经网络(x=12.95%)和SVM模型(x=3.19%),表明ELM模型更加精确、高效。     

Analysis of the relationship between ambient levels of O3, NO2 and NO as a function of NOx in the UK

Monitoring data from the UK Automatic Urban and Rural Network are used to investigate the relationships between ambient levels of ozone (O₃), nitric oxide (NO) and nitrogen dioxide (NO₂) as a function of NO_x, for levels ranging from those typical of UK rural sites to those observed at polluted urban kerbside sites. Particular emphasis is placed on establishing how the level of ‘oxidant’, OX (taken to be the sum of O₃ and NO₂) varies with the level of NO_x, and therefore to gain some insight into the atmospheric sources of OX, particularly at polluted urban locations. The analyses indicate that the level of OX at a given location is made up of NO_x-independent and NO_x-dependent contributions. The former is effectively a regional contribution which equates to the regional background O₃ level, whereas the latter is effectively a local contribution which correlates with the level of primary pollution. The local oxidant source has probable contributions from (i) direct NO₂ emissions, (ii) the thermal reaction of NO with O₂ at high NO_x, and (iii) common-source emission of species which promote NO to NO₂ conversion. The final category may include nitrous acid (HONO), which appears to be emitted directly in vehicle exhaust, and is potentially photolysed to generate HO_x radicals on a short timescale throughout the year at southern UK latitudes. The analyses also show that the local oxidant source has significant site-to-site variations, and possible reasons for these variations are discussed. Relationships between OX and NO_x, based on annual mean data, and fitted functions describing the relative contributions to OX made by NO₂ and O₃, are used to define expressions which describe the likely variation of annual mean NO₂ as a function of NO_x at 14 urban and suburban sites, and which can take account of possible changes in the regional background of O₃.     

人工神经网络和专家系统在污水生物处理系统中的应用

对近年来国内外污水生物处理系统中人工神经网络和专家系统的应用进行了简要的回顾。分析了废水生物处理工艺难于控制的原因及人工神经网络和专家系统的结构和特点。结果表明．国外智能控制发展迅速,并且应用领域遍及污水生物处理的各个方面,国内尚处于起步阶段。简要探讨了废水生物处理智能控制今后应深入研究的问题及方向。     

Neural network forecasting of air pollutants hourly concentrations using optimised temporal averages of meteorological variables and pollutant concentrations

The new method for the forecasting hourly concentrations of air pollutants is presented in the paper. The method was developed for a site in urban residential area in city of Zagreb, Croatia, for four air pollutants (NO₂, O₃, CO and PM₁₀). Meteorological variables and concentrations of the respective pollutant were taken as predictors. A novel approach, based on families of univariate regression models, was employed in selecting the averaging intervals for input variables. For each variable and each averaging period between 1 and 97 h, a separate model was built. By inspecting values of the coefficient of correlation between measured and modelled concentrations, optimal averaging periods for each variable were selected. A new dataset for building the forecasting model was then calculated as temporal moving averages (running means) of former variables. A multi-layer perceptron type of neural networks is used as the forecasting model. Index of agreement, calculated for the entire dataset including the data for model building, ranged from 0.91 to 0.97 for the respective pollutants. As suggested by the analysis of the relative importance of the input variables, different agreements for different pollutants are likely due to different sources and production mechanisms of investigated pollutants. A comparison of the new method with more traditional method, which takes hourly averages of the forecast hour as input variables, showed similar or better performance. The model was developed for the purpose of public-health-oriented air quality forecasting, aiming to use a numerical weather forecast model for the prediction of the part of input data yet unknown at the forecasting time. It is to expect that longer term averages used as inputs in the proposed method will contribute to smaller input errors and the greater accuracy of the model.     

融合拓扑信息与气象信息的空气质量预测网络

通过空气质量监测数据对正在形成或即将到来的空气污染进行预测是一项具有重要意义的工作,而空气质量监测站只能检测其周围一定范围内的空气污染情况。为了衡量整个城市的空气污染情况,获取任意时间、任意位置的空气质量信息,结合交叉注意力机制,提出了一种融合拓扑信息与气象信息的空气质量预测网络(CGMIM)。将西安市空气质量监测数据与气象数据转换为图像拼接起来,作为输入信息。在高阶非线性时空动态神经网络(MIM)的基础上引入注意力机制,并增加拓扑图编码器模块,提高模型提取能力以及对空气质量监测数据中的空间特征的利用率。最后,使用时空损失函数替代传统的均方误差损失函数,提高模型对空间关系的关注。结果表明：CGMIM网络模型能够在准确预测的同时,对位置区域合理填充,能够有效提升空气质量监测数据的空间分辨率。     

Short-term Model of the Production of Construction Aggregates in Taiwan Based on Artificial Neural Networks

BACKGROUND: Taiwan's geography and limited stock of sandstone have caused sandstone resources to gradually decline to the point of exhaustion after long-term excavation. Moreover, the Taiwanese government has continuously increased the amount of land area near rivers that cannot be excavated to facilitate riverbed remediation and promote conservation of water resources. Accordingly, predicting and managing the annual production of construction aggregates in future construction projects, and dealing appropriately with some thorny problems, for instance, demand that excess supply, excessive excavation, unregulated excavation, and the consequent environmental damage, will significantly affect the efficient use of natural resources in a manner that accords with the national policy of Sustainable Development (SD). METHODS:. This study establishes an empirical model for forecasting the annual production of future construction aggregates using Artificial Neural Networks (ANN), based on 15 relevant socio-economic indicators, such as indicator of annual consumption of cement. A sensitivity analysis is then performed on these indicators. RESULTS AND DISCUSSION: This work applies ANN to estimate the annual production of construction aggregates; the estimates, the verification of the model and the sensitivity analysis are all acceptable. Furthermore, sensitivity analysis results indicate that the annual consumption of cement is the indicator that most strongly influences the production of construction aggregates, as well as whether construction waste can be recycled and steel structures can be used in buildings, helping to reduce the future production of construction aggregates in Taiwan. CONCLUSIONS: The elaborate prediction methodology presented in this study avoids some of the weaknesses or limitations of conventional linear statistics, linear programming or system dynamics. Additionally, the results not only provide a short-term prediction of the production of construction aggregates in Taiwan, but also provide a viable and flexible means of verifying quality certification of the production data of construction aggregates in the future by incorporating those relevant socio-economic indicators. RECOMMENDATIONS AND OUTLOOK: The continuity and quality of the database of relevant indicators used in this study should be closely scrutinized in order to ensure the SD means of exploiting resources.     

Implications of network particle tracking (NPT) for ecological model interpretation

Network particle tracking (NPT), building on the foundation of network environ analysis (NEA), is a new development in the definition of coherence relations within and between connected systems. This paper evaluates three ecosystem models in a comparison of throughflow- and storage-based NEA and NPT. Compartments in models with high indirect effects and Finn cycling showed low correlation of NEA storage and throughflow with particle repeat visits and numbers of particles in compartments at steady state. Conversely, the correlation between NEA and NPT results was high with two models having lower indirect effects and Finn cycling. Analysis of ecological orientors associated with NEA showed NPT to fully support conventional NEA results when the common conditions of donor control and steady state are satisfied. Particle trajectories are recorded in the new concept of a particle “passport”. Ability to track and record particle in-system histories enables views of multiple scales and opens the possibility of making pathway-dependent modeling decisions. NPT may also enable modeling of time, allowing integration of Newtonian, organismal and stochastic modeling perspectives in a single comprehensive analysis.