A knowledge-based network for Section Yidong Bridge,Dongyang River,one tributary of Qiantang River,Zhejiang Province,China,is established in order to model water quality in areas under small data.Then,based on normal ...A knowledge-based network for Section Yidong Bridge,Dongyang River,one tributary of Qiantang River,Zhejiang Province,China,is established in order to model water quality in areas under small data.Then,based on normal transformation of variables with routine monitoring data and normal assumption of variables without routine monitoring data,a conditional linear Gaussian Bayesian network is constructed.A "two-constraint selection" procedure is proposed to estimate potential parameter values under small data.Among all potential parameter values,the ones that are most probable are selected as the "representatives".Finally,the risks of pollutant concentration exceeding national water quality standards are calculated and pollution reduction decisions for decision-making reference are proposed.The final results show that conditional linear Gaussian Bayesian network and "two-constraint selection" procedure are very useful in evaluating risks when there is limited data and can help managers to make sound decisions under small data.展开更多
The U.S. EPA （Environmental Protection Agency） established the CASTNET （Clean Air Status and Trends Network） and its predecessor, the NDDN （national dry deposition network）, as national air quality and meteorolo...The U.S. EPA （Environmental Protection Agency） established the CASTNET （Clean Air Status and Trends Network） and its predecessor, the NDDN （national dry deposition network）, as national air quality and meteorological monitoring networks. Both CASTNET and NDDN were designed to measure concentrations of sulfur and nitrogen gases and particles. Both networks also estimate dry deposition using an inferential model. The design was based on the concept that atmospheric dry deposition flux could be estimated as the product of a measured air pollutant concentration and a modeled deposition velocity （Vd）. The MLM （multi-layer model）, the computer model used to simulate dry deposition, requires information on meteorological conditions and vegetative cover as model input. The MLM calculates hourly Fa for each pollutant, but any missing meteorological data for an hour renders Vd missing for that hour. Because of percent completeness requirements for aggregating data for long-term estimates, annual deposition rates for some sites are not always available primarily because of missing or invalid meteorological input data. In this work, three methods for replacing missing on-site measurements are investigated. These include （1） using historical values of deposition velocity or （2） historical meteorological measurements from the site being modeled or （3） current meteorological data from nearby sites to substitute for missing inputs and thereby improve data completeness for the network＇s dry deposition estimates. Results for a CASTNET site used to test the methods show promise for using historical measurements of weekly average meteorological parameters.展开更多
In recent years, the bridge safety monitoring has been paid more attention in engineering field. How- ever, the financial and material resources as well as human resources were costly for the traditional monitoring me...In recent years, the bridge safety monitoring has been paid more attention in engineering field. How- ever, the financial and material resources as well as human resources were costly for the traditional monitoring means. Besides, the traditional means of monitoring were low in accuracy. From an engineering example, based on neural network method and historical data of the bridge monitoring to construct the BP neural network model with dual hidden layer strueture, the bridge temperature field and its effect on the behavior of bridge deflection are forecasted. The fact indicates that the predicted biggest error is 3.06% of the bridge temperature field and the bridge deflection behavior under temperature field affected is 2. 17% by the method of the BP neural net-work, which fully meet the precision requirements of the construction with practical value.展开更多
基金Project(50809058)supported by the National Natural Science Foundation of China
文摘A knowledge-based network for Section Yidong Bridge,Dongyang River,one tributary of Qiantang River,Zhejiang Province,China,is established in order to model water quality in areas under small data.Then,based on normal transformation of variables with routine monitoring data and normal assumption of variables without routine monitoring data,a conditional linear Gaussian Bayesian network is constructed.A "two-constraint selection" procedure is proposed to estimate potential parameter values under small data.Among all potential parameter values,the ones that are most probable are selected as the "representatives".Finally,the risks of pollutant concentration exceeding national water quality standards are calculated and pollution reduction decisions for decision-making reference are proposed.The final results show that conditional linear Gaussian Bayesian network and "two-constraint selection" procedure are very useful in evaluating risks when there is limited data and can help managers to make sound decisions under small data.
文摘The U.S. EPA （Environmental Protection Agency） established the CASTNET （Clean Air Status and Trends Network） and its predecessor, the NDDN （national dry deposition network）, as national air quality and meteorological monitoring networks. Both CASTNET and NDDN were designed to measure concentrations of sulfur and nitrogen gases and particles. Both networks also estimate dry deposition using an inferential model. The design was based on the concept that atmospheric dry deposition flux could be estimated as the product of a measured air pollutant concentration and a modeled deposition velocity （Vd）. The MLM （multi-layer model）, the computer model used to simulate dry deposition, requires information on meteorological conditions and vegetative cover as model input. The MLM calculates hourly Fa for each pollutant, but any missing meteorological data for an hour renders Vd missing for that hour. Because of percent completeness requirements for aggregating data for long-term estimates, annual deposition rates for some sites are not always available primarily because of missing or invalid meteorological input data. In this work, three methods for replacing missing on-site measurements are investigated. These include （1） using historical values of deposition velocity or （2） historical meteorological measurements from the site being modeled or （3） current meteorological data from nearby sites to substitute for missing inputs and thereby improve data completeness for the network＇s dry deposition estimates. Results for a CASTNET site used to test the methods show promise for using historical measurements of weekly average meteorological parameters.
文摘In recent years, the bridge safety monitoring has been paid more attention in engineering field. How- ever, the financial and material resources as well as human resources were costly for the traditional monitoring means. Besides, the traditional means of monitoring were low in accuracy. From an engineering example, based on neural network method and historical data of the bridge monitoring to construct the BP neural network model with dual hidden layer strueture, the bridge temperature field and its effect on the behavior of bridge deflection are forecasted. The fact indicates that the predicted biggest error is 3.06% of the bridge temperature field and the bridge deflection behavior under temperature field affected is 2. 17% by the method of the BP neural net-work, which fully meet the precision requirements of the construction with practical value.
