基于MFOA-SVR露天矿边坡变形量预测研究

为实现边坡危险性及时预警预报,以露天矿边坡变形量为研究对象,提出采用七项影响指标作为边坡位移变形量的响应参数,建立支持向量机回归预测模型(SVR)。引入修正的果蝇优化算法(MFOA)对模型参数进行优化,构建基于MFOA-SVR露天矿边坡变形量协同预测模型,并以实际监测数据进行模型仿真预测。结果表明:该模型平均绝对误差为0.9167mm,平均相对误差为4.2737%,较其他模型预测精度高,综合性能好,将其运用于露天矿边坡变形量预测研究具有较好的适用性和可靠性。     

基于粗糙集和支持向量机的标准农田地力等级评价

标准农田是耕地的精华,是确保国家粮食安全的关键。科学评价标准农田地力等级对标准农田培肥和土壤改良有着重要意义。将粗糙集(Rough Set,RS)理论和支持向量机(Support Vector Machine,SVM)相结合,提出了基于RS和SVM的标准农田地力等级评价方法,同时,利用遗传算法的并行搜索结构和模拟退火的概率突跳特性,提出了GASA优化SVM参数算法。该方法首先在确定标准农田地力等级评价指标的基础上,利用地力调查样本数据及传统的指数和法评价结果构建RS决策表,应用RS穷尽算法对决策表进行约简,剔除冗余的评价指标,然后用约简后的评价指标作为SVM的输入,运用GASA优化SVM参数算法对SVM进行训练,建立标准农田地力等级的RS-SVM评价模型。应用该方法对温州市鹿城区标准农田地力等级进行评价,与未用RS约简的SVM模型和BP神经网络模型评价结果进行对比,SVM模型和BP神经网络模型的输入指标数均为15个,其评价正确率分别为100%和90%;RS-SVM模型的输入指标数为14个,其评价正确率分别为100%,结果表明,该方法通过RS约简评价指标后,SVM评价精度并没有降低,但降低了SVM输入向量维数和计算复杂度,提高了训练效率;SVM 用于标准农田地力等级评价,具有比BP神经网络更高的评价精度,可有效用于标准农田地力等级评价,为耕地地力评价提供了新方法。     

Sustainability assessment of entire forest value chains: Integrating stakeholder perspectives and indicators in decision support tools

The optimization of value chains is an important process to promote sustainable development, since value chains are closely linked to the satisfaction of human needs and combine different driving forces for environmental change. This article presents a methodological approach for the participatory development of value-chain wide sustainability indicator sets and their integration into a decision support tool in the specific case study of the chain “construction and refurbishment with wood”. There are numerous indicator sets for sustainable development of forests and sustainable forestry available at different levels, ranging from local, regional and national to global scale assessments. Some efforts were also made to integrate later production stages of forest value chains (such as wood processing) in the assessment scope (e.g. for chain-of-custody certification). However, no indicator set has so far been available covering environmental, social and economic aspects for the entire value chain of building with timber. This gap was closed through applied sustainability research in the project “Holzwende 2020: Sustainable future markets for wood in the building sector”.     

民用飞行器重着陆事件决策系统研究

应用支持向量决策方法,解决飞行器重着陆事件诊断中受力分析复杂、样本量小的问题。探索飞行器接地阶段重着陆事件的形成机制,分别建立翼根载荷与起落架载荷的受力模型,研究飞机主体载荷与重着陆事件的非线性关系。研究重着陆决策所需样本集的小样本特性,提出支撑向量驱动的智能决策规则方法,进行重着陆事件诊断模型的建模研究。最后,建立重着陆事件决策系统。实例表明:该决策方法对于实例样本的诊断精度达到95.7%,对于提高着陆阶段的安全具有实际意义。     

Linking linear programming and spatial simulation models to predict landscape effects of forest management alternatives

Forest management planners require analytical tools to assess the effects of alternative strategies on the sometimes disparate benefits from forests such as timber production and wildlife habitat. We assessed the spatial patterns of alternative management strategies by linking two models that were developed for different purposes. We used a linear programming model (Spectrum) to optimize timber harvest schedules, then a simulation model (HARVEST) to project those schedules in a spatially explicit way and produce maps from which the spatial pattern of habitat could be calculated. We demonstrated the power of this approach by evaluating alternative plans developed for a national forest plan revision in Wisconsin, USA. The amount of forest interior habitat was inversely related to the amount of timber cut, and increased under the alternatives compared to the current plan. The amount of edge habitat was positively related to the amount of timber cut, and increased under all alternatives. The amount of mature northern hardwood interior and edge habitat increased for all alternatives, but mature pine habitat area varied. Mature age classes of all forest types increased, and young classes decreased under all alternatives. The average size of patches (defined by age class) generally decreased. These results are consistent with the design goals of each of the alternatives, but reveal that the spatial differences among the alternatives are modest. These complementary models are valuable for quantifying and comparing the spatial effects of alternative management strategies.     

地铁深基坑内支撑结构优化分析

深基坑工程中不同支撑数量会影响基坑系统支护结构的刚度,将对基坑的变形产生一定的影响.武汉市某地铁车站深基坑工程在设计前期分别提出了两种基坑支护结构设计方案,一种采用五道内支撑,另一种采用六道内支撑.为验证不同内支撑数量对该深基坑支护结构的影响,采用Plaxis 3D有限元软件建立了三维有限元模型,模拟分析了不同内支撑数...     

科技支撑美丽中国建设的进展和展望

自党的十八大提出美丽中国建设目标以来,为发挥好科技支撑作用,国家组织开展了一系列关于资源、生态和环境等领域的科技计划,有效支撑了美丽中国建设的理论探索、生态环境保护修复和生态文明体制建设决策支持等方面的科技需求。本文在梳理过去十年国际及发达国家生态环境领域科技研发布局情况,以及我国相关领域科技研发计划和推进情况的基础上,重点介绍了中国科学院“美丽中国生态文明建设科技工程”战略性先导科技专项（A类）的主要研发内容及取得的阶段进展,并基于现状与期望分析,辨识有关领域科技前沿动态与发展趋势,提出关于下一步科技发展方向的建议,以期为党的二十大之后科技推进美丽中国建设的方向提供参考。     

面向驾驶员辅助系统的换道意图辨识方法研究

为提高先进驾驶员辅助系统工作效能,通过差异性分析选取了人-车-路特征参数,建立了支持向量机换道意图辨识模型.基于驾驶模拟器采集的数据,运用ROC曲线对比分析了不同参数组合、时窗、数据表达形式下模型的分类效果,结果表明以方向盘转角、加速度、车辆距车道线距离和驾驶人头动信息为特征参数,时窗为1.5s,以时窗内信息均值和方差表述特征参量可获得最优辨识效果.研究表明,模型在保证1 s前瞻性的同时辨识率可达93％.     

Study on urban water environmental support capacity

ＳｔｕｄｙｏｎｕｒｂａｎｗａｔｅｒｅｎｖｉｒｏｎｍｅｎｔａｌｓｕｐｐｏｒｔｃａｐａｃｉｔｙＧｕｏＨｕａｉｃｈｅｎｇ；ＹｅＷｅｎｈｕ（ＣｅｎｔｅｒｏｆＥｎｖｉｒｏｎｍｅｎｔａｌＳｃｉｅｎｃｅｓ，ＰｅｋｉｎｇＵｎｉｖｅｒｓｉｔｙ，Ｂｅｉｊｉｎｇ１００８７...     

Landscape fragmentation by urbanization in the Netherlands:optionsand ecological consequences

１Ｉｎｔｒｏｄｕｃｔｉｏｎ：ｔｈｅｓｔｕｄｙａｒｅａＦｉｇ１ＴｈｅｓｔｕｄｙａｒｅａＴｈｅｅｃｏｎｏｍｉｃａｃｔｉｖｉｔｉｅｓｏｆｔｈｅＮｅｔｈｅｒｌａｎｄｓａｒｅｃｏｎｃｅｎｔｒａｔｅｄｉｎｔｈｅｗｅｓｔｏｆｔｈｅｃｏｕｎｔｒｙ，ｔｈｅｓｏｃａ...