In?ltration percolation is a rustic and extensive purifying technique which is capable of completely oxidizing and decontaminating wastewater. The object of the present study is to validate on a real scale the finding...In?ltration percolation is a rustic and extensive purifying technique which is capable of completely oxidizing and decontaminating wastewater. The object of the present study is to validate on a real scale the findings gotten in pilot laboratory. It is notably to confirming the purification performances and decontamination of secondary effluent by sand filter drained under an intense wild plant. For this, a basin of 100 m2 of infiltration surface equipped with two wells of sampling in the center and in the periphery, has been constructed in irrigated perimeter of Dissa in Gabes (south east of Tunisia). Our results show that essential of the COD is eliminated in first 50 centimeters in sand filter and the presence of wild plant on the surface of infiltration damaged quality of filtered water. The follow-up of the organic nitrogen and ammonium show the good performances of process in elimination of nitrogenous pollution. The rate of exhaustion is about 100% and this since first 50 centimeters of filtration. The sand filter retains the different shapes of the phosphor by adsorption and precipitation in the superficial horizon. The elimination of the bacterial pollution is high, it varying from 0.1 to 2.5 Ulog. The bacterial purification remained low after 50 cm filtration. Indeed, with 50 cm depth, the quality of the filtrate is slightly better than the influent with still tendency to higher removal in the center than in periphery of the basin.展开更多
对比仅包含多光谱信息、仅可实现二维土地覆盖分类的传统光学遥感数据,机载多光谱激光雷达(multispectral light detection and ranging,MS-LiDAR)的优势在于同时包含多光谱和空间信息、可实现三维土地覆盖分类,但现有的机载MS-LiDAR数...对比仅包含多光谱信息、仅可实现二维土地覆盖分类的传统光学遥感数据,机载多光谱激光雷达(multispectral light detection and ranging,MS-LiDAR)的优势在于同时包含多光谱和空间信息、可实现三维土地覆盖分类,但现有的机载MS-LiDAR数据的土地覆盖分类研究所需特征维度过高、算法复杂度高。因此,提出了一种整合空间相关性和归一化差分比率指数(Normalized Difference Ratio Index,NDRI)特征的逐步分类算法。该算法首先融合机载MS-LiDAR数据的多波段独立点云,获取兼具空间位置及其多光谱信息的单一点云数据;然后利用空间邻域增长下的地面滤波算法分离地面和非地面点;接着基于不同目标的激光反射特性差异设计将草地(树木)自地面(非地面)中分离的NDRI指数,并利用类间方差最大原则下的自适应最优NDRI指数实现地面和非地面点的精细分类;最后利用3D多数投票法优化分类结果。采用加拿大Optech Titan实测MS-LiDAR数据测试提出算法的有效性及可行性,实验结果表明:算法的平均总体精度和Kappa系数分别可达90.17%和0.861,可有效实现城区MS-LiDAR数据的三维土地覆盖分类;分步处理的方式更有利于针对具体的分离目标的特点设计简单且有效的规则,算法设计更简单、复杂度低;NDRI可为其他机器学习算法的显著性特征的设计和选择提供理论支撑。展开更多
随着互联网上Web服务的日益增多,面对大量功能相同的候选服务,用户希望能够选择质量最优的候选服务.然而,用户通常并不知道所有候选服务的服务质量(Quality of Service,QoS).因此,基于Web服务的历史记录预测QoS值得到了广泛关注.传统的...随着互联网上Web服务的日益增多,面对大量功能相同的候选服务,用户希望能够选择质量最优的候选服务.然而,用户通常并不知道所有候选服务的服务质量(Quality of Service,QoS).因此,基于Web服务的历史记录预测QoS值得到了广泛关注.传统的基于协同过滤(CF)的预测方法可能会遭遇数据稀疏、用户信任等问题,导致该方法在预测精度方面表现一般.为解决上述问题,该文提出一种基于覆盖随机游走算法的服务质量预测方法.该方法首先基于用户服务历史QoS记录,使用改进的覆盖算法对用户进行聚类,选取与每个用户聚类次数的Top-k个用户为该用户的信任用户,连接所有用户与其信任用户构建用户信任网;其次,基于用户信任网提出一种随机游走预测方法,在随机游走的过程中,不仅考虑目标服务的QoS信息,同时考虑相似服务的QoS信息,以确保QoS预测的准确性;最后,每次随机游走获得一个QoS值,为使预测更加准确,作者进行多次随机游走,汇总所有QoS值进行预测.为验证文中方法的有效性,作者在真实的Web服务数据集进行了大量实验,其中包括来自339个用户的5825个真实世界Web服务的1 974 675个Web服务调用.实验结果表明文中方法在预测精度上明显优于现有方法,同时可以很好地解决推荐系统的数据稀疏和用户信任问题.展开更多
文摘In?ltration percolation is a rustic and extensive purifying technique which is capable of completely oxidizing and decontaminating wastewater. The object of the present study is to validate on a real scale the findings gotten in pilot laboratory. It is notably to confirming the purification performances and decontamination of secondary effluent by sand filter drained under an intense wild plant. For this, a basin of 100 m2 of infiltration surface equipped with two wells of sampling in the center and in the periphery, has been constructed in irrigated perimeter of Dissa in Gabes (south east of Tunisia). Our results show that essential of the COD is eliminated in first 50 centimeters in sand filter and the presence of wild plant on the surface of infiltration damaged quality of filtered water. The follow-up of the organic nitrogen and ammonium show the good performances of process in elimination of nitrogenous pollution. The rate of exhaustion is about 100% and this since first 50 centimeters of filtration. The sand filter retains the different shapes of the phosphor by adsorption and precipitation in the superficial horizon. The elimination of the bacterial pollution is high, it varying from 0.1 to 2.5 Ulog. The bacterial purification remained low after 50 cm filtration. Indeed, with 50 cm depth, the quality of the filtrate is slightly better than the influent with still tendency to higher removal in the center than in periphery of the basin.
文摘对比仅包含多光谱信息、仅可实现二维土地覆盖分类的传统光学遥感数据,机载多光谱激光雷达(multispectral light detection and ranging,MS-LiDAR)的优势在于同时包含多光谱和空间信息、可实现三维土地覆盖分类,但现有的机载MS-LiDAR数据的土地覆盖分类研究所需特征维度过高、算法复杂度高。因此,提出了一种整合空间相关性和归一化差分比率指数(Normalized Difference Ratio Index,NDRI)特征的逐步分类算法。该算法首先融合机载MS-LiDAR数据的多波段独立点云,获取兼具空间位置及其多光谱信息的单一点云数据;然后利用空间邻域增长下的地面滤波算法分离地面和非地面点;接着基于不同目标的激光反射特性差异设计将草地(树木)自地面(非地面)中分离的NDRI指数,并利用类间方差最大原则下的自适应最优NDRI指数实现地面和非地面点的精细分类;最后利用3D多数投票法优化分类结果。采用加拿大Optech Titan实测MS-LiDAR数据测试提出算法的有效性及可行性,实验结果表明:算法的平均总体精度和Kappa系数分别可达90.17%和0.861,可有效实现城区MS-LiDAR数据的三维土地覆盖分类;分步处理的方式更有利于针对具体的分离目标的特点设计简单且有效的规则,算法设计更简单、复杂度低;NDRI可为其他机器学习算法的显著性特征的设计和选择提供理论支撑。
文摘随着互联网上Web服务的日益增多,面对大量功能相同的候选服务,用户希望能够选择质量最优的候选服务.然而,用户通常并不知道所有候选服务的服务质量(Quality of Service,QoS).因此,基于Web服务的历史记录预测QoS值得到了广泛关注.传统的基于协同过滤(CF)的预测方法可能会遭遇数据稀疏、用户信任等问题,导致该方法在预测精度方面表现一般.为解决上述问题,该文提出一种基于覆盖随机游走算法的服务质量预测方法.该方法首先基于用户服务历史QoS记录,使用改进的覆盖算法对用户进行聚类,选取与每个用户聚类次数的Top-k个用户为该用户的信任用户,连接所有用户与其信任用户构建用户信任网;其次,基于用户信任网提出一种随机游走预测方法,在随机游走的过程中,不仅考虑目标服务的QoS信息,同时考虑相似服务的QoS信息,以确保QoS预测的准确性;最后,每次随机游走获得一个QoS值,为使预测更加准确,作者进行多次随机游走,汇总所有QoS值进行预测.为验证文中方法的有效性,作者在真实的Web服务数据集进行了大量实验,其中包括来自339个用户的5825个真实世界Web服务的1 974 675个Web服务调用.实验结果表明文中方法在预测精度上明显优于现有方法,同时可以很好地解决推荐系统的数据稀疏和用户信任问题.
