基于模糊C均值聚类和随机森林的短时交通状态预测方法

交通拥堵长期以来是城市面临的主要问题之一,解决交通拥堵瓶颈刻不容缓。准确的短时交通状态预测有利于市民预知交通出行信息,及时采取措施避免陷入拥堵困境。该文提出一种基于模糊C均值聚类(FCM)和随机森林的短时交通状态预测方法。首先,利用一种新颖的融合时空信息的自适应多核支持向量机(AMSVM)来预测短时交通流参数,包括流量、速度和占有率。其次,基于FCM算法分析历史交通流,获取历史交通状态信息。最后,利用随机森林算法分析所预测的短时交通流参数,得到最终预测的短时交通状态。该方法在融合时空信息的同时采用随机森林算法应用于短时交通状态预测这一全新的研究领域。实验结果表明,FCM对历史交通状态的评估方式适用于不同的高速路和城市道路场景。其次,随机森林比其它常见的机器学习方法具有更高的预测精度,从而提供实时可靠的短时交通出行信息。     

火线轮廓参数卫星遥感定量提取方法

对Landsat TM/ETM+数据进行抽样统计分析基础上,把Landsat TM/ETM+的红外、短波红外和近红外等波段数据相结合,采用窗口动态阈值算法构建燃烧区识别模型;在此基础上,为定量生成火线轮廓参数,通过连通性判断、孔洞填充、小斑块去除和边缘细化等图像处理方法对识别的燃烧区进行处理;并在ENVI 4.8+IDL语言环境下,实现了基于Landsat TM/ETM+数据自动生成火线轮廓参数算法处理过程的程序化.结果表明,总体判对率为86.44%,总体误判率为13.56%(其中漏判率为1.77%,错判率为11.79%);该方法可满足林火扑救中对火线轮廓参数定量宏观监测的应用需求.     

森林资源亚健康状态的卫星遥感预警技术研究

森林病虫害等扰动类型造成的亚健康林木监测预警工作不能及时到位,导致防治工作长期处于灾后救灾的被动局面.基于2019年5～9月份的多时相GF-1 WFV数据,应用比值植被指数和红绿植被指数,准实时地监测逆生长、叶冠胁迫或失色等"灾害"信息.结果 表明:虽然树木叶片枯黄、萎蔫等叶绿素降解并逐渐转化成叶黄素和叶红素需要一定的...     

基于表示学习和深度森林的长链非编码RNA编码短肽预测模型

长链非编码RNA（lncRNA）中的小开放阅读框（sORFs）能够编码长度不超过100个氨基酸的短肽。针对短肽预测研究中lncRNA中的sORFs特征不鲜明且高可信度数据尚不充分的问题,提出一种基于表示学习的深度森林（DF）模型。首先,使用常规lncRNA特征提取方法对sORFs进行编码;其次,通过自编码器（AE）进行表示学习来获得输入数据的高效表示;最后,训练DF模型实现对lncRNA编码短肽的预测。实验结果表明,该模型在拟南芥数据集上能够达到92.08%的准确率,高于传统机器学习模型、深度学习模型以及组合模型,且具有较好的稳定性;此外,在大豆与玉米数据集上进行的模型测试中,该模型的准确率分别能达到78.16%和74.92%,验证了所提模型良好的泛化能力。     

Estimating aboveground carbon in a catchment of the Siberian forest tundra: Combining satellite imagery and field inventory

This study was part of an interdisciplinary research project on soil carbon and phytomass dynamics of boreal and arctic permafrost landscapes. The 45 ha study area was a catchment located in the forest tundra in northern Siberia, approximately 100 km north of the Arctic Circle.The objective of this study was to estimate aboveground carbon (AGC) and assess and model its spatial variability. We combined multi-spectral high resolution remote sensing imagery and sample based field inventory data by means of the k-nearest neighbor (k-NN) technique and linear regression.Field data was collected by stratified systematic sampling in August 2006 with a total sample size of n = 31 circular nested sample plots of 154 m² for trees and shrubs and 1 m² for ground vegetation. Destructive biomass samples were taken on a sub-sample for fresh weight and moisture content. Species-specific allometric biomass models were constructed to predict dry biomass from diameter at breast height (dbh) for trees and from elliptic projection areas for shrubs.Quickbird data (standard imagery product), acquired shortly before the field campaign and archived ASTER data (Level-1B product) of 2001 were geo-referenced, converted to calibrated radiances at sensor and used as carrier data. Spectral information of the pixels which were located in the inventory plots were extracted and analyzed as reference set. Stepwise multiple linear regression was applied to identify suitable predictors from the set of variables of the original satellite bands, vegetation indices and texture metrics. To produce thematic carbon maps, carbon values were predicted for all pixels of the investigated satellite scenes. For this prediction, we compared the kNN distance-weighted classifier and multiple linear regression with respect to their predictions.The estimated mean value of aboveground carbon from stratified sampling in the field is 15.3 t/ha (standard error SE = 1.50 t/ha, SE% = 9.8%). Zonal prediction from the k-NN method for the Quickbird image as carrier is 14.7 t/ha with a root mean square error RMSE = 6.42 t/ha, RMSE_r = 44%) resulting from leave-one-out cross-validation. The k-NN-approach allows mapping and analysis of the spatial variability of AGC. The results show high spatial variability with AGC predictions ranging from 4.3 t/ha to 28.8 t/ha, reflecting the highly heterogeneous conditions in those permafrost-influenced landscapes. The means and totals of linear regression and k-NN predictions revealed only small differences but some regional distinctions were recognized in the maps.     

Metrics for evaluating the forest biorefinery supply chain performance

For sustainable decision-making regarding biorefinery strategies, different criteria, i.e. economic, environmental, social, should be considered. However, the economic criteria typically do not consider market volatility, whereas today's market involves price and demand volatilities. Biorefinery strategies must be flexible to be robust to market volatility. Therefore, relevant metrics must be developed to quantify the system's performance against volatility. This paper presents metrics of flexibility and robustness which analyze the performance of the supply chain in a dynamic environment, providing additional information along with economic metrics. In this paper, the link between the two metrics, and how profitability and robustness change with flexibility are discussed. The results reveal that, although profitability does not always increase with more flexibility and there is an optimum level of flexibility, the system's robustness is improved by increasing flexibility. Moreover, a “conditional value-at-risk” parameter is introduced to show what patterns of sale lead to highest profit and robustnestness.     

Segment-constrained regression tree estimation of forest stand height from very high spatial resolution panchromatic imagery over a boreal environment

Mean stand height is an important parameter for forest volume and biomass estimation in support of monitoring and management activities. Information on mean stand height is typically obtained through the manual interpretation of aerial photography, often supplemented by the collection of field calibration data. In remote areas where forest management practices may not be spatially exhaustive or where it is difficult to acquire aerial photography, alternate approaches for estimating stand height are required. One approach is to use very high spatial resolution (VHSR) satellite imagery (pixels sided less than 1 m) as a surrogate for air photos. In this research we demonstrate an approach for modelling mean stand height at four sites in the Yukon Territory, Canada, from QuickBird panchromatic imagery. An object-based approach was used to generate homogenous segments from the imagery (analogous to manually delineated forest stands) and an algorithm was used to automatically delineate individual tree crowns within the segments. A regression tree was used to predict mean stand height from stand-level metrics generated from the image grey-levels and within-stand objects relating individual tree crown characteristics. Heights were manually interpreted from the QuickBird imagery and divided into separate sets of calibration and validation data. The effects of calibration data set size and the input metrics used on the regression tree results were also assessed. The approach resulted in a model with a significant R² of 0.53 and an RMSE of 2.84 m. In addition, 84.6% of the stand height estimates were within the acceptable error for photo interpreted heights, as specified by the forest inventory standards of British Columbia. Furthermore, residual errors from the model were smallest for the stands that had larger mean heights (i.e., > 20 m), which aids in reducing error in subsequent estimates of biomass or volume (since stands with larger trees contribute more to overall estimates of volume or biomass). Estimated and manually interpreted heights were reclassified into 5-metre height classes (a schema frequently used for forest analysis and modelling applications) and compared; classes corresponded in 54% of stands assessed, and all stands had an estimated height class that was within ± 1 class of their actual class. This study demonstrates the capacity of VHSR panchromatic imagery (in this case QuickBird) for generating useful estimates of mean stand heights in unmonitored, remote, or inaccessible forest areas.     

ArcFVS:An integration of a Geographic Information System and Forest Vegetation Simulator

This paper presents the design and implementation of an integrated application program called ArcFVS that links the Forest Vegetation Simulator (FVS) and a Geographic Information System (GIS) to realize spatial selection of input files and graphic display of modeling output. Data for testing and running the model came from the U.S. Forest Service's Forest Inventory and Analysis (FIA) Database and were also collected in field surveys in north-central Indiana. ArcFVS 1.0 is designed using the ArcGIS software from the Environmental Systems Research Institute (ESRI) and the Visual Basic for Applications (VBA) programming environment to manipulate ESRI's ArcObjects. The resulting product offers custom functions as commands in a new menu or as tools on a new toolbar. They are used to: select spatially or by attribute the forest plots to be projected by FVS, create the FVS input files for the selected plots and display in a geospatial environment different types of FVS output (text output files, tables with variables of interest and visualization image files). Advantages of ArcFVS 1.0 include the new GIS capabilities, enhanced format translation functions and the standardized programming environment.     

黑龙江大兴安岭新林-塔河金成矿潜力分析

大兴安岭地区采金历史悠久,砂金资源十分丰富。岩金找矿以来,经过水系沉积物测量、土壤地球化学测量等方法,发现了一批金矿床或矿点。结合已往工作及区域成矿条件,以地质信息为基础,对区内找矿潜力进行分析。以化探信息为先导,地、物、化、遥成矿信息综合标志确定预测远景区的界线,共划分A类远景区2处,B类远景区2处,C类远景区3处,勘查靶区3处,成矿潜力较大,为下步区域找矿提供思路。     

鸿恩寺森林公园鸿恩阁夜景照明设计——地域性照明设计手法探索

通过对国内传统风格建筑夜景照明现状的分析,针对传统风格建筑夜景照明中出现的现实问题,以西南地区仿古建筑的夜景照明实践探索为例,以地域性为切入点,探讨如何使传统风格建筑照明设计适应具体的空间地域要求。