Plenty of inorganic nitrogen in wastewater can cause the eutrophication in water bodies, so it is an important task to remove nitrogen. Purification role was realized by absorption, filtration, depositon, evaporation,...Plenty of inorganic nitrogen in wastewater can cause the eutrophication in water bodies, so it is an important task to remove nitrogen. Purification role was realized by absorption, filtration, depositon, evaporation, nitrification and denitrification of microbes. Although the studies of Phragmites austrilis bed in the constructed wetland are popular, the purification performances of constructed wetland filled by saline-alkali soil substrate are less reported. In the paper, the purification efficiency of nitrogen with Phragmites austrilis bed in the constructed wetland filled by saline-alkali soil substrate was discussed through a simulation study. Results to date indicated that the first order plug flow model was adequate to describe the nitrogen removal. The experiment showed that the diminishing concentration of TN, NO 2-N, NO 3-N, NH 4-N were closely related to hydrological retention time (HRT), the correlation coefficient was R 2=0.98499, R 2=0.9911, R 2=0.89407 and R 2=0.95459, respectively. According to the data, the most suitable hydrological retention time (HRT) for this kind of constructed wetland should be determined to 4 days. In addition, the experiment showed the purification efficiency of nitrogen has very broad range and drastic vibration, TN (17%-79%), NO 2-N (33%-98%), NO 3-N(13%-93%), NH 4-N (28%-64%). The study will promote wetland’s design and operation procedures in large saline-alkaline soil areas.展开更多
A new class of support vector machine, nu-support vector machine, is discussed which can handle both classification and regression. We focus on nu-support vector machine regression and use it for phase space predictio...A new class of support vector machine, nu-support vector machine, is discussed which can handle both classification and regression. We focus on nu-support vector machine regression and use it for phase space prediction of compares nu-support vector machine with back propagation (BP) networks in order to better evaluate the performance of the proposed methods. The experimental results show that the nu-support vector machine regression obtains lower root mean squared error than the BP networks and provides an accurate chaotic time series prediction. These results can be attributable to the fact that nu-support vector machine implements the structural risk minimization principle and this leads to better generalization than the BP networks.展开更多
基金NortheastInstituteofGeographyAgriculturalEcology,CAS .(KZCX3 SW NA 13 )
文摘Plenty of inorganic nitrogen in wastewater can cause the eutrophication in water bodies, so it is an important task to remove nitrogen. Purification role was realized by absorption, filtration, depositon, evaporation, nitrification and denitrification of microbes. Although the studies of Phragmites austrilis bed in the constructed wetland are popular, the purification performances of constructed wetland filled by saline-alkali soil substrate are less reported. In the paper, the purification efficiency of nitrogen with Phragmites austrilis bed in the constructed wetland filled by saline-alkali soil substrate was discussed through a simulation study. Results to date indicated that the first order plug flow model was adequate to describe the nitrogen removal. The experiment showed that the diminishing concentration of TN, NO 2-N, NO 3-N, NH 4-N were closely related to hydrological retention time (HRT), the correlation coefficient was R 2=0.98499, R 2=0.9911, R 2=0.89407 and R 2=0.95459, respectively. According to the data, the most suitable hydrological retention time (HRT) for this kind of constructed wetland should be determined to 4 days. In addition, the experiment showed the purification efficiency of nitrogen has very broad range and drastic vibration, TN (17%-79%), NO 2-N (33%-98%), NO 3-N(13%-93%), NH 4-N (28%-64%). The study will promote wetland’s design and operation procedures in large saline-alkaline soil areas.
文摘A new class of support vector machine, nu-support vector machine, is discussed which can handle both classification and regression. We focus on nu-support vector machine regression and use it for phase space prediction of compares nu-support vector machine with back propagation (BP) networks in order to better evaluate the performance of the proposed methods. The experimental results show that the nu-support vector machine regression obtains lower root mean squared error than the BP networks and provides an accurate chaotic time series prediction. These results can be attributable to the fact that nu-support vector machine implements the structural risk minimization principle and this leads to better generalization than the BP networks.