新型FDR土壤水盐一体传感器标定研究与应用

运用新型频域反射(FDR)技术和分频技术,实现传感器水盐同时测量。因土壤质地、容重等差异,土壤水盐标定是利用FDR监测土壤水盐过程中的必要环节。以东北壤土为试验对象,利用"由湿到干,由小到大"的土壤水盐运动规律,通过土柱试验,用烘干法和土壤溶液电导率法对FDR进行室内标定研究。同时,将时域反射(TDR)式水盐传感器作为对照系,进行稳定性、精准度比较。结果表明:1)经土壤水分标定曲线校正后,FDR式水盐一体传感器具有很高的测量精度,可达±1%~3%;2)在土壤里对同一水盐含量进行百次重复测定,TDR传感器水分平均值为20.07%,相对标准偏差为0.18%;盐分(电导率)平均值为0.29 m S/cm,相对标准偏差为11.06%;新型FDR传感器水分平均值为20.08%,相对标准差为0.21%;盐分(电导率)平均值为0.31m S/cm,相对标准差为8.49%。新型FDR土壤水盐一体传感器可用于农田土壤连续水盐监测,提供稳定、精确的基础数据,结合数据分析软件,指导种植生产。     

基于水分驱动的AquaCrop模型及其研究进展

介绍了AquaCrop模型的原理及基本参数,从模型的校验与应用两方面阐述了该模型的研究进展。指出目前仍缺乏实测数据验证AquaCrop模型对蒸发及蒸腾的模拟效果;AquaCrop模型在严重水分及盐分胁迫下模拟结果精度较差;已开展的模拟研究地域范围窄;由于缺少更复杂的生理子模块,AquaCrop模型不能很好解释水分胁迫对光合产物向籽粒运输分配过程的影响。为了提高模型的模拟精度并进一步延伸模型的应用范围,应完善模型水分及盐胁迫模块,并在较广范围内获取丰富的实测数据对模型开展进一步的校验研究。     

基于宽域氧传感器的发动机空燃比测试系统设计

为了实现发动机电子控制系统对空燃比的精确反馈控制和方便对发动机进行诊断等用途,在熟悉宽域氧传感器工作原理的基础上,设计了空燃比采集显示系统和PC监控系统。并对测试系统的输出电压和混合气的过量空气系数的对应关系进行了试验标定,最后在发动机试验台架上该测试系统进行了试验验证。结果表明该系统灵敏度高、响应时间短（100 ms）,采集数据显示直观,可用于发动机空燃比闭环控制和车载自动诊断系统（OBD）等的开发。     

近红外光谱法测定玉米秸秆饲用品质

为了对玉米秸秆的饲用品质进行可靠、便捷、快速的分析和评价,该研究以不同品种、密度、氮肥和水分处理的不同发育时期和不同部位玉米秸秆为试验材料,应用近红外光谱（NIRS）技术和偏最小二乘法（PLS）,采用一阶导数+中心化+多元散射校正的光谱数据预处理方法,构建了玉米秸秆体外干物质消化率（IVDMD）、酸性洗涤纤维（ADF）、中性洗涤纤维（NDF） 和可溶性糖（WSC）含量的NIRS分析模型。所建立的IVDMD、ADF、NDF和WSC含量的NIRS校正模型决定系数（R2cal）分别为0.9906、0.9870、0.9931和0.9802,交叉验证决定系数（R2cv）分别为0.9593、0.9413 、0.9678和0.9342,外部验证决定系数（R2val）分别为0.9549、0.9353、0.9519和0.9191,各项标准差（SEC、SECV和SEP）为0.935～1.904,相对分析误差（RPD）均大于3。结果表明,各参数的NIRS分析模型可用于玉米秸秆饲用品质的分析和品种选育的快速鉴定。     

基于垄线平行特征的视觉导航多垄线识别

为有效快速地识别农田多条垄线以实现农业机器人视觉导航与定位,提出一种基于机器视觉的田间多垄线识别与定位方法。使用VC++ 6.0开发了农业机器人视觉导航定位图像处理软件。该方法通过图像预处理获得各垄行所在区域,使用垂直投影法提取出导航定位点。根据摄像机标定原理与透视变换原理,计算出各导航定位点世界坐标。然后结合垄线基本平行的特征,使用改进的基于Hough变换的农田多垄线识别算法,实现多垄线的识别与定位。使用多幅农田图像进行试验并在室内进行了模拟试验。处理一幅320×240的农田图像约耗时219.4 ms,室内试验各垄线导航距与导航角的平均误差分别为2.33 mm与0.3°。结果表明,该方法能有效识别与定位农田的多条垄线,同时算法的实时性也能满足 要求。     

立体视觉技术在森林资源调查中的应用

将立体视觉应用于森林资源调查是一种全新的手段 ,能高效地对样地进行调查 ,并且为高分辨率遥感图像提供一种获取林分调查因子的技术 .该文从林业环境的实际情况出发 ,探索立体视觉技术应用于森林资源调查的可行技术路线 .针对树干边缘的特点 ,提出了Canny算子检测的树干边缘和外极线约束下、交互式的区域相关立体匹配策略 .采用在中国林业科学研究院院内树林中FUJIFILMFinePix 6 90 0Zoom数码相机拍摄的立体图像对 ,实现了立体视觉技术测量树木点位和胸径 ,并估计了在实验条件下的量测精度 .     

Applicability of ground‐penetrating radar as a tool for nondestructive soil‐depth mapping on Pleistocene periglacial slope deposits

Soil depth plays a decisive role in determining soil properties in mountainous regions for ecological site assessment. To evaluate the use of ground‐penetrating radar (GPR) for fast and high‐resolution mapping within mountainous regions, we examined the possibilities and limitations of GPR to determine soil depth over bedrock and to delineate individual substrate layers formed during the Pleistocene in a periglacial environment (Pleistocene periglacial slope deposits, PPSD). Selected catenae in representative subregions of the study area (Dill catchment, SE Rhenish Massif, Germany) have been successfully mapped using GPR. A practicable method was developed using a 400 MHz antenna to reach a mean penetration depth of 1.5 m and to map different substrates and layers of PPSD based on calibrations of the GPR at soil pits along 12 catenae. Colluvium, the three types of PPSD layers, as well as the in situ bedrock could be distinguished in most sections of the GPR surveys. Characteristic GPR facies caused by intrinsic material properties of the different substrates, such as stone content and soil moisture content, could be distinguished in different geomorphologic and lithological settings. A layer‐based velocity distribution was determined for characteristic substrate layers at soil pits enabling us to considerably enhance the accuracy of soil‐depth prediction. Compared to traditionally surveyed soil profiles, our results demonstrate an accuracy of layer thickness surveying within a standard deviation of approx. 0.1 m. It is demonstrated that the combination of GPR with conventional soil‐pit mapping is an efficient and valid method to produce high‐resolution data of substrate distribution.     

Soil‐carbon turnover under different crop management: Evaluation of RothC‐model predictions under Pannonian climate conditions

Despite the importance of soil organic matter (SOM), very few long‐term data concerning soil organic‐C dynamics are available for calibrating and evaluating C models. The long‐term ¹⁴C‐turnover field experiment, established in 1967 in Fuchsenbigl, Lower Austria, offers the unique opportunity to follow the fate of labeled C under different crop‐management systems (bare fallow, spring wheat, crop rotation) over a period of more than 35 y. Compared with the crop‐rotation and spring‐wheat treatments, the decline of total organic C was largest in the bare‐fallow treatments, because no significant C input has occurred since 1967. Nonetheless, the decline was not as fast as predicted with the original RothC‐26.3‐model decomposition rate constants. In this work, we therefore calibrated the Roth‐C‐26.3 model for the Pannonian climatic region based on the field‐experiment results. The main adjustment was in the decomposition rate constant for the humified soil C pool (HUM), which was set to 0.009 instead of 0.02 y^–1 as determined in the original Rothamsted field trial. This resulted in a higher HUM pool in the calibrated model because of a longer turnover period (111 vs. 50 y). The modeled output based on the calibrated model fitted better to measured values than output obtained with the original Roth‐C‐26.3‐model parameters. Additionally, the original decomposition rate constant for resistant plant material (RPM) was changed from 0.3 to 0.6 y^–1 to describe the decomposition of ¹⁴C‐labeled straw more accurately. Application of the calibrated model (modified HUM decomposition rate) to simulate removal of crop residues showed that this can entail a long‐term decline of SOM. However, these impacts are strongly dependent on the crop types and on environmental conditions at a given location.     

中子仪测定土壤水分方法的研究进展

中子仪测定土壤水分方法的核心内容是中子计数与土壤含水量的曲线标定,为了绘制精确的能够应用于不同土壤层次、土壤质地、土壤有机质含量和土壤剖面含水量变化等条件的标定曲线,通过对已取得的不同土壤条件下标定曲线文献的整理分析,从中子仪测水的工作原理出发,分析了影响中子仪测定土壤水分的主要因素,概括了测水所遇到的主要问题及其解决的方法,并对准确测定含水量层状变化土壤,提出标定曲线绘制的可行方法。根据中子计数值不仅与土壤容积含水量有关,而且还与水分和中子计数器的距离有关,提出了进一步准确标定曲线的研究方法。     

近红外透射光谱技术测定黍稷蛋白含量的研究

为了探索快速测定完整黍稷籽粒蛋白含量的方法,采用近红外光谱分析技术建立数学模型并进行预测,比较原始透射光谱经导数处理结合不同回归算法对模型的影响。结果表明,分别经一阶和二阶导数处理后利用偏最小二乘法和改进的偏最小二乘法,4种方法的分析效果相近,最优的是一阶导数结合改进的偏最小二乘回归法,黍稷蛋白定标模型的定标相关系数（RSQ）为0.8806,定标标准误差（SEC）为0.3424,交互验证相关系数（1-VR）为0.8570,交互验证标准误差（SECV）为0.3751,外部预测标准偏差（SEP）为0.454。最终以完整黍稷籽粒为样品所建立的蛋白NITS模型,可以用于黍稷蛋白含量的快速检测。