Using improved neural network model to analyze RSP,NOx and NO2 levels in urban air in Mong Kok,Hong Kong

As the health impact of air pollutants existing in ambient addresses much attention in recent years, forecasting of airpollutant parameters becomes an important and popular topic inenvironmental science. Airborne pollution is a serious, and willbe a major problem in Hong Kong within the next few years. InHong Kong, Respirable Suspended Particulate (RSP) and NitrogenOxides NO_x and NO₂ are major air pollutants due to thedominant diesel fuel usage by public transportation and heavyvehicles. Hence, the investigation and prediction of the influence and the tendency of these pollutants are ofsignificance to public and the city image. The multi-layerperceptron (MLP) neural network is regarded as a reliable andcost-effective method to achieve such tasks. The works presentedhere involve developing an improved neural network model, whichcombines the principal component analysis (PCA) technique and theradial basis function (RBF) network, and forecasting thepollutant levels and tendencies based in the recorded data. Inthe study, the PCA is firstly used to reduce and orthogonalizethe original input variables (data), these treated variables arethen used as new input vectors in RBF neural network modelestablished for forecasting the pollutant tendencies. Comparingwith the general neural network models, the proposed modelpossesses simpler network architecture, faster training speed,and more satisfactory predicting performance. This improvedmodel is evaluated by using hourly time series of RSP, NO_x and NO₂ concentrations collected at Mong Kok Roadside Gaseous Monitory Station in Hong Kong during the year 2000. By comparing the predicted RSP, NO_x and NO₂ concentrationswith the actual data of these pollutants recorded at the monitorystation, the effectiveness of the proposed model has been proven.Therefore, in authors' opinion, the model presented in the paper is a potential tool in forecasting air quality parameters and hasadvantages over the traditional neural network methods.     

基于B-P神经网络的环境质量评价方法

提出可将环境质量评价的无论是定量指标还是定性参数转化成"二进制"的"1"或"0",进而将这种二进制数引入B-P网络.通过实例探讨,这种新的B-P网络既适用于定量指标的水质参数又适用于定性指标的水质参数.     

Ambient Air Quality Monitoring Network Design for Assessing Human Health Impacts from Exposures to Airborne Contaminants

Existing methods of establishing ambientair quality monitoring networks typically evaluateonly parameters related to ambient concentrations ofthe contaminant(s) of interest such as emissionsource characteristics, atmospheric transport anddispersion, secondary reactions, depositioncharacteristics, and local topography. However,adverse health risks from exposures to airbornecontaminants are a function of the contaminant andthe anatomic and physiologic characteristics of theexposed population. Thus, ambient air qualitymonitoring networks designed for the protection ofpublic health or for epidemiological studiesevaluating adverse health impacts from exposures toambient air contaminants should account for bothcontaminant characteristics and human healthparameters. A methodology has been establishedwhich optimizes ambient air quality monitoringnetworks for assessments of adverse human healthimpacts from exposures to airborne contaminants byincorporating human health risk assessmenttechniques. The use of risk assessment techniquesas the basis for designing ambient air qualitymonitoring networks will help to target limitedfinancial and human resources to evaluate humanhealth risks from exposures to airbornecontaminants.     

Determining Ecoregions for Environmental and GMO Monitoring Networks

A representative environmental monitoring network at the regional scale cannot use raster-based or random sampling designs, but requires a stratified sampling procedure integrating different information layers, and it has to occur in ecologically differing homogeneous regions (ecoregions). These we have determined using a set of spatial strata with ecological variables which we analysed with classification and regression trees (CART). We present a framework for environmental monitoring, that covers different scales, and we transfer the framework to a potential GMO (genetically modified organisms) monitoring network. We use ecoregion and other environmental strata together with existing environmental monitoring networks to determine GMO monitoring sites more precisely.     

Land User Participation in Developing a Computerised Decision Support System for Combating Desertification

Combating desertification in natural rangelands has recently become a priority in large parts of southern Africa. Rangeland managers, farmers, scientists, conservationists and land users have been applying a variety of restoration technologies to address this problem. Bush encroachment, as part of the desertification process, involves the natural replacement of the herbaceous plant cover by undesirable problem woody species. The active and passive restoration technologies that are applied, are mainly based on indigenous knowledge and include the chemical, mechanical or manual reclamation of unproductive rangelands, as well as the combating of woody and alien species encroachment. Indigenous practices and knowledge play a major role in the effectiveness and success rate of these technologies. This project faces the challenge of bringing together both local and scientific knowledge in a single user-friendly, computerised Decision Support System (DSS) which is directly accessible by land users to support them in the process of decision making, concerning the combating of desertification. Case studies from central and northern Namibia were used to combine qualitative and quantitative data to develop this Decision Support System. The DSS currently consists of two databases and an expert system, which evaluates the results of land users’ management practices, and provides easily accessible information and advice for participants in the system, based on the incorporated data. The DSS is also linked to national and international web sites and databases to offer a wider range of information on technologies concerning agricultural and conservation practices.     

人工神经网络用于铅的化学形态模拟计算

用前馈线性网络法求解水体系中Ｐｂ（２＋）与ＯＨ－之间的反应常数，不同训练算法对求解结果的精度、收敛速度及权值均有影响．结果表明，批处理算法的精度最好，权值不出现负值，但运算时间最长；在线算法的精度虽不如批处理算法，而比数据变换－在线算法好，权值有时会出现负值．运算时间较长；数据变换－在线算法的优点是运算时间短，但相对误差较大，权值出现负值的机会多。采用反馈网络模拟计算铅的各种化学形态的浓度．用物料核算的方法对反馈网络模型进行检验表明，此种模型用于平衡计算是可行的，详细分析了理论模拟和实验曲线的差异的原因，温度的影响最小，在４＜ｐＨ＜９时，ＣＯ有重要的影响．在国代检验时，ｎ值取整所引入的误差的影响亦不可忽视。从本文的结果可以看到，采用前馈网络和反馈网络相结合的方法考察水体中的化学形态是可行的．从而为解决这一类问题提供了一种可能的途径．     

基团贡献法对取代苯类化合物生物降解性的预测

测定了 47种取代苯类化合物在松花江水中的 5日生化需氧量 (BOD5) .分别采用线性基团贡献法和非线性基团贡献法(人工神经网络法 )对化合物的生物降解性BOD5 ThOD(ThOD :理论需氧量 )进行QSBR研究 .得到不同基团对生物降解性的贡献为 :C6H5>COOH >OH >CH3 O CH3 >NH2 >Cl >NO2 .线性基团贡献法对于训练组和测试组的定性预测正确率分别为72 %和 86 % ;而人工神经网络法的预测正确率分别为 92 %和 86 % .预测结果表明线性和非线性基团贡献法的预测效果都很好 ,相比而言 ,非线性方法对生物降解性的预测更准确     

Detecting neural tube defects by amniocentesis between 11 and 15 weeks' gestation

Forty-two open neural tube defects (NTDs) were identified in our series of 7440 amniocenteses tested between 11 and 15 weeks of gestation. Using a cut-off of ≥2.0 MOM, the detection rate for open NTDs was 95 per cent; 100 per cent each for anencephaly and spina bifida; and 78 per cent for encephalocele. Two encephaloceles had AFP levels less than 2.0 MOM and negative AChEs. Thirty-four (81 per cent) of these NTDs were tested between 13 and 15 weeks and 8 (19 per cent) before 13 weeks. There were 0.6 per cent false positives by AFP (excluding serious abnormalities and fetal death) and 0.1 per cent after AChE. The likelihood of an open NTD after an elevated AFP (≥2.0 MOM) was 24 and 77 per cent for any serious abnormality. These results, when combined with an earlier study, indicate that amniotic fluid AFP appears to be as sensitive a test for open NTDs between 13 and 15 weeks as between 16 and 20 weeks. Additional experience is necessary to determine this before 13 weeks.     

利用神经网络及数值模拟获取变压边力控制曲线

在应用Dynaform、Autoform等板料数值模拟软件的过程中 ,通常需要对压边力施力曲线进行各种预先的设定 ,但如何选定施力模式一直是一个比较棘手的问题 ,在文中利用数值模拟及MATLAB中的神经网络工具箱得出了压边力的最优控制曲线     

BP人工神经网络预测邻苯二甲酸酯光化学降解

将反向传播人工神经网络（BP－ANN）用于邻苯二甲酸酯类化合物的光化学降解的预测中，选取正交设计的试验点作为反向传播人工神经网络的训练集，实现对全实验域试验点的预测，并与实测的试验数据比较。结果表明反向传播人工神经网络（BP－ANN）具有良好的预测能力。