Effect of tile effluent on nutrient concentration and retention efficiency in agricultural drainage ditches

Tile drainage is a common water management practice in many agricultural landscapes in the Midwestern United States. Drainage ditches regularly receive water from agricultural fields through these tile drains. This field-scale study was conducted to determine the impact of tile discharge on ambient nutrient concentration, nutrient retention and transport in drainage ditches. Grab water samples were collected during three flow regimes for the determination of soluble phosphorus (SP), ammonium nitrogen (NH₄⁺-N), nitrate nitrogen (NO₃-N) concentrations and their retention in three drainage ditches. Measured nutrient concentration indicated lower SP and NH₄⁺-N, and greater NO₃-N concentrations in tile effluents compared to the ditch water. Net uptake lengths were relatively long, especially for NO₃-N, indicating that nutrients were generally not assimilated efficiently in these drainage systems. Results also indicated that the study reaches were very dynamic showing alternating increases or decreases in nutrient concentration across the flow regimes. The drainage ditches appeared to be nutrient-rich streams that could potentially influence the quality of downstream waters.     

Optimized denitrification bioreactor treatment through simulated drainage containment

In the design of wood-based, enhanced-denitrification bioreactors to treat nitrate in agricultural drainage, the consideration of the highly variable flow rates and nitrate concentrations inherent to many drainage systems is important. For optimized mitigation of these nitrate loads, it may be best to contain drainage water prior to treatment in order to facilitate longer, more constant retention times rather than to allow cycles of flushing and dry periods in the denitrification bioreactor. Simulated containment prior to bioreactor treatment compared to passing drainage directly through a bioreactor was investigated with the use of six pilot-scale denitrification bioreactors constructed with plywood and filled with Pinus radiata woodchips at Massey University No. 4 Dairy Farm (Palmerston North, New Zealand). Initial bromide tracer tests were followed with a series of five simulated drainage events each at successively declining inflow nitrate concentrations. During each drainage event, three pilot bioreactors received a simulated hydrograph lasting 1.5 days (Non-Containment treatment) and three pilot bioreactors received the same total drainage volume treated over 4 days at a constant flow rate (i.e. constant retention time; Containment treatment). Results showed significantly different total mass removal efficiencies of 14.0% vs. 36.9% and significantly different removal rates of 2.1 g N m⁻³ day⁻¹ vs. 6.7 g N m⁻³ day⁻¹ for the Non-Containment and Containment treatments, respectively, which indicated that treating drainage at constant retention times provided more optimized nitrate removal. While this work was done to evaluate treatment under New Zealand drainage conditions, it also provides valuable information for optimizing agricultural drainage denitrification bioreactor performance in general.     

达拉特旗井灌条件下农田水利用及地下水动态分析研究

以达拉特旗壕庆河流域为研究对象,通过对该区地下水位和水质的监测及农田水利用状况的调查,摸清地下水的年动态变化规律及水质分布,初步估算该研究区农田灌溉地下水年开采量,从而为本流域的地下水资源评价及合理开发,高效利用提供一定的依据。根据数据分析结果,地下水是从西南流向东北方向,补给黄河南岸引黄灌区。观测期间地下水位年际变化受降雨的影响较大,总体呈上升趋势,变化幅度为0.5～1.5m。pH值超过8.5,呈碱性水,部分区域地下水矿化度为0.5～0.9g/L,可用于农业灌溉,随着地下水流动的方向,水质逐渐变差,下游的盐分浓度最大到2.6g/L,需要合理进行灌溉防止土壤盐碱化。通过遥感图片识别分析计算出该区农田面积为1 398hm2,农业灌溉年开采地下水量为516万m3。     

县域农机化技术推广工作探讨

在新的形势下,县域农机推广部门肩负着认真做好技术推广工作的重任,农机推广部门应该明确目标,履行职责,抓住重点,强化自身素质,更快更好地做好农机化普及工作.     

基于粒子滤波的苹果采摘机器人目标姿态估算

针对苹果采摘机器人中果实姿态信息缺失造成果实损伤及采摘失败等问题,根据苹果尾部花萼遗迹区域是否可见分析了果实姿态信息机器视觉测量方法,研究了果实姿态信息的粒子滤波估计。将单目摄像机固定在机器人末端手爪上,在手爪趋近果实的采摘过程中,随机采集多帧果实图像。对每帧图像运用惯性主轴或重心点偏移两种方法计算果实在机器人坐标系下的姿态向量值,并通过粒子滤波融合来得到果实姿态信息的最优估计。实验结果表明,粒子滤波方法可以有效地融合多帧图像的苹果姿态信息,能够减小因依靠单帧图像直接测量而产生的误差。     

基于水平集和先验信息的农业图像分割方法

提出了一种基于先验信息的C-V模型并对杂草、小麦、苹果进行分割研究.根据某类农业图像的特点,把图像表示为易于分割的模型,提取模型中感兴趣目标的信息量作为先验信息,通过H分量得到初始轮廓,并以此初始化提出的模型,迭代求解水平集函数,得到收敛的目标轮廓曲线.对杂草、小麦、苹果分割结果统计分割面积正确率为0.999、0.999、0.846,面积错误率为0、0、0.125.     

人工神经网络NIR定量分析方法及其软件实现

在Visual C＋＋环境中采用面向对象技术，开发了PCA-MBP-NIR定量分析模型软件。通过40份小麦样品的原始光谱、加噪光谱（信噪比为14dB）与含水率所建立的PLS-NIR与PCA-MBP-NIR模型，对10份未知小麦样品的原始光谱、加噪光谱分别进行含水率的PLS-NIR与PCA-MBP-NIR预测分析。分析表明，对于含噪声的光谱，与PLS建模相比，使用PCA-MBP-NIR对未知样品预测结果具有更高的相关系数，更低的预测误差标准差。     

小型农产品分选机设计与试验

针对家庭农场及小规模营销的需要,设计了基于机器视觉的小型农产品分选机。农产品在平胶带上形成多通道阵列式输送,以DSP作为机器视觉单元的核心,采集一帧图像可分析多个通道,并根据农产品的形状、颜色等图像特征进行分选。在水平输送速度为9.2 cm/s、农产品输送间隔为16 cm、4通道并行输送的条件下,选用核桃、红枣及栗子3种农产品进行了分选机性能测试。试验表明,机器能够可靠工作;核桃、红枣和栗子的分选准确性分别为91.66%、94.79%和97.39%;分选速度达8 800个/h。因此采用DSP芯片作为小型分选机的机器视觉核心是可行的,可以为农产品分选机的小型化、低成本提供技术支持。     

基于Harris算子的籽粒尖端识别方法

通过籽粒尖端形态特征分析,提出了一种基于Harris角点检测的籽粒尖端识别方法。Harris算子在一个局部区域中检测角特征,作为籽粒最明显的角结构,尖端在Harris角点检测中具有最大的响应值,使其对尖端不明显的籽粒也有良好的响应。通过对玉米、南瓜和西葫芦等具有尖端特征的750粒籽粒的测试结果表明,提出算法的尖端检测综合准确率为95.6%。     

农业机器人自主导航改进自适应滤波控制器研究

为提高导航精度与控制精度,对农业机器人的自主导航控制进行研究。结合Sage-Husa自适应滤波与强跟踪卡尔曼滤波两种算法优点,利用严格收敛判据,设计了改进的自适应卡尔曼滤波算法。新算法保证了系统的实时性和稳定性,且具有更好的滤波精度。利用变结构切换方法来阻止PID控制器积分饱和现象,有效解决了控制器过饱和问题。将改进的自适应滤波算法与变结构PID控制器相结合,可提高导航系统的稳定性和精度。对所提理论进行仿真与试验,结果表明该方法大幅度提高了滤波器抑制发散的能力和导航的控制精度。