SPOT5卫星影像在退耕精确监测中的应用研究

退耕还林(草)是针对水土流失治理、改善西部脆弱的生态环境而采取的一项重大举措。以2003年9月9日获取的庄浪县SPOT5卫星影像为例,讨论了SPOT5卫星影像的正射纠正方法,详细介绍了影像纠正前根据参考图选点误差、数字高程模型的误差精度和传感器内部结合误差三方面对影像纠正应该达到的精度理论值进行评估,以及纠正后对影像几何精度的精度评定的方法。讨论了在缺少蓝色波段的情况下通过波段计算合成真彩色影像的技术和处理方法,并结合退耕还林工程的实际情况进一步探讨了经过处理后的SPOT5影像在退耕信息可视化和精确监测中的应用。     

The evolution of Greenways in China

This paper focuses on the evolution of Greenway planning and implementation in China, and provides a historical context to the Greenway concept. It was found that:

(1) Although the concept of Greenway was an adaptation from the Western World, the Chinese have a history of more than 2000 years of Greenway planning and implementation. Chinese Greenways have been called various names and were planned for various reasons.

(2) The long history of Greenway planning and implementation in China was mainly a “top-down” approach, which, while very effective under a centralized administrative system, often lacked a scientific basis and significant public participation.

(3) The functions of the Greenways were mainly protection and productive, with little concern for human uses such recreational uses of cycling and hiking.

Greenways in China are discussed chronologically and in three categories:

(1) Riparian Greenways run along rivers, streams and water channels. The history of these Greenways dates more than 2000 years, since a time when trees were grown along canals and city moats. They have in modern times evolved into a network of drainage channels.

(2) Greenways along transportation corridors. These Greenways run parallel to state and provincial highways, railroads, country roads and urban streets and evolved from tree plantings along highways. Used exclusively by emperors, the green corridor networks have been systematically planned and constructed at a national scale and directly organized by the central government.

(3) Greenways along farmland for wind protection. These plantings evolved from individual segments of windbreak rows to a network of protective windbreaks and the large, regional scale “Green Great Wall” project running along the northern edge of China.

As Greenways have evolved in China, they reflect changes in ideology, utilization and scale; from protection of production or beautification to ecological and multiple uses, and from small-scale fragments to a systematic regional and national network. The occurrence of disasters, the involvement of state leaders and the influence of science played an important role in the evolution of Greenways in China.The paper also argues that:

• the traditional top-down approach in Greenway planning and implement should be integrated with scientifically based methods;

• recreational uses should be considered and integrated into existing and planned Greenways;

• the recently invoked “city beautiful”, or cosmetic approach to Greenway planning and implementation should be stopped; and

• Greenways should be planned as an critical strategic element of ecological infrastructure at both the regional and urban scale during current rapid and extensive urbanization occurring in China.

基于新的森林优化算法的特征选择算法

针对传统的基于森林优化算法的特征选择算法在初始化阶段、候选森林生成阶段和更新阶段存在的问题,提出了一种新的基于森林优化算法的特征选择算法。该算法在初始化阶段采用皮尔森相关系数和L1正则化方法代替随机初始化策略;在候选森林生成阶段,采用优劣树分开和差额补足的方法解决优劣树不完备问题;在更新阶段,将与最优树精度相同但维度不同的树木添加到森林中。在实验中,所提算法采用与传统的基于森林优化算法的特征选择算法相同的实验数据和实验参数,分别测试了小维度、中维度和大维度数据。实验结果表明,在2个大维度数据和2个中维度数据上,所提算法的分类精度和维度缩减能力均高于传统的基于森林优化算法的特征选择算法。实验结果验证了所提算法在处理特征选择问题的有效性。     

基于随机森林与变邻域下降的车辆合乘求解

为了最大化用户满意度，长期车辆合乘问题（LTCPP）被建模为多目标优化问题。然后，根据历史合乘数据以及用户满意度信息，使用随机森林算法计算每个指标对用户满意度的重要性影响，并作为对应优化目标的权重，以避免人为设定权重因子对优化结果的影响。提出了一种求解LTCPP的变邻域下降（VND）算法，通过顺序地在多个邻域内搜索得到问题的最优解。实验结果表明，结合随机森林和VND算法能为LTCPP提供高质量的解决方案，且具有很高的时间效率。     

基于随机森林的变压器油中溶解气体浓度预测

对油中溶解气体浓度进行分析及发展趋势预测,可以为变压器的状态评估提供重要的依据。传统的离线DGA方法因易导致延迟判断变压器的运行状态,造成一定的经济损失,现已不适用于油中溶解气体浓度分析及预测。因此,提出一种基于随机森林的变压器油中溶解气体浓度预测模型,以更准确地分析与预测油中溶解气体浓度。该模型以7种气体浓度构成特征向量空间,作为可视输入,并以目标气体浓度作为输出。试验结果表明,相较于传统的机器学习方法(BPNN、RBF和SVM),随机森林模型能更准确地预测油中溶解气体浓度,且需要调整参数少、训练效率高。通过算例分析,验证了该方法的有效性。     

基于RF-GA的六维力传感器解耦方法

多维力传感器的维间耦合问题严重影响了检测精度的提高。通过设计新型RF-GA（基于遗传算法的改进随机森林算法）解耦方法解决多维力信息的解耦问题，实现提高力传感器检测精度的目标。针对随机森林算法中含有大量子树，但每个子树的预测准度无法保证的问题，利用遗传算法对随机森林的子树进行筛选，保留优质子树，从而提高预测精度。以基于应变检测的六维力传感器为实验对象，将RF-GA算法运用到实际力信息解耦中，并通过解耦实验对RF-GA算法进行验证。与现有解耦算法相比，RF-GA解耦方法具有精度高、解耦时间短的优点，实验结果表明该算法能有效提高多维力传感器的解耦精度。     

北京市新城滨河森林公园建设标准探索

对北京市新城滨河森林公园实施方案评审工作中所探索出的建设标准进行了总结，形成了以乔木种植为主，提高常绿苗木比例，提高种植密度，注重林水结合，统一规划，根据实际需求分批建设公园基础设施和服务设施等城市滨河森林公园建设的新理念。     

Stakeholder perspectives on converting forest biomass to energy in Oregon, USA

Within the state of Oregon, USA, there is considerable interest in the possibility of converting forest biomass to energy. A number of studies have assessed the technical feasibility of forest biomass energy, but few have focused on social aspects, an important consideration in projects involving public forests. This study explores the social context of converting forest biomass to energy, using qualitative research methods. Semi-structured interviews were conducted with forty individuals representing nine different stakeholder groups. Information gained through interviews was used to understand stakeholder views on forest biomass energy, including their perspectives on potential barriers and opportunities in Oregon. Findings indicate the most challenging barrier will be access to long-term, consistent supply. A related challenge is the long history of contention between parties over forest products coming from public lands. However, findings also show that there are many areas of common ground between these groups that have historically been at odds, such as agreement on the necessity of restoration treatments in certain forest types, the by-product of which could be used for biomass generation. Potential conflicts still exist, for instance over projects in mixed conifer forests. Development of policies and projects through inclusive, collaborative approaches could alleviate controversies, potentially allowing more activities to move forward. Information provided by this research creates a foundation for discussions as forest biomass energy becomes an increasingly prominent issue in Oregon, the western USA, and other regions of the world.     

崇义客家梯田区森林枯落物持水特性

为了解崇义客家梯田水源区森林枯落物水文特性,进一步揭示森林的水源涵养功能,通过现场调查取样和室内实验分析等方法,对崇义客家梯田区森林枯落物的储量和持水率、持水量、吸水速率等持水特性进行了研究。结果表明:试验地各林分类型枯落物总厚度为3.0~8.6 cm,总储量为3.29~13.19 t/hm²,总储量整体上随总厚度的增加而增大;枯落物最大持水率为239.65%~343.54%,即枯落物最大吸持水量可达其干重的2.4~3.4倍,枯落物最大持水量为5.58~12.32 t/hm²,相当于0.56~1.23 mm的降雨水深;在浸泡开始2 h内,枯落物吸持水量增加较快,随着浸泡时间延长,持水量与吸水速率的变化均较为平缓,持水量与浸泡时间呈明显对数函数关系,吸水速率与浸泡时间呈明显幂函数关系。研究结果可为区域林分类型调配、森林涵养水源功能监测提供科学依据。     

SPT based determination of undrained shear strength: Regression models and machine learning

The purpose of this study is the accurate prediction of undrained shear strength using Standard Penetration Test results and soil consistency indices, such as water content and Atterberg limits. With this study, along with the conventional methods of simple and multiple linear regression models, three machine learning algorithms, random forest, gradient boosting and stacked models, are developed for prediction of undrained shear strength. These models are employed on a relatively large data set from different projects around Turkey covering 230 observations. As an improvement over the available studies in literature, this study utilizes correct statistical analyses techniques on a relatively large database, such as using a train/test split on the data set to avoid overfitting of the developed models. Furthermore, the validity and consistency of the prediction results are ensured with the correct use of statistical measures like p-value and cross-validation which were missing in previous studies. To compare the performances of the models developed in this study with the prior ones existing in literature, all models were applied on the test data set and their performances are evaluated in terms of the resulting root mean squared error (RMSE) values and coefficient of determination (R²). Accordingly, the models developed in this study demonstrate superior prediction capabilities compared to all of the prior studies. Moreover, to facilitate the use of machine learning algorithms for prediction purposes, entire source code prepared for this study and the collected data set are provided as supplements of this study.