Dietary selenate versus selenite for cattle, sheep, and horses.

Food and Drug Administration regulations currently permit addition of .3 mg of Se per kilogram of diet for chickens, turkeys, ducks, swine, sheep, and cattle. However, field reports indicate that this level may not be adequate for ruminants in all situations. Because sodium selenite is the most common supplemental form and is known to be readily absorbed to particles or reduced to insoluble elemental Se or selenides in acid, anaerobic environments, studies were conducted with dairy cattle, sheep, and horses fed sodium selenate to determine whether Se from this source was more bioavailable than Se from sodium selenite. A 2-wk period of no Se supplementation was followed by 49 or 56 d of Se supplementation at .3 mg/kg of dietary DM. Serum Se concentrations and glutathione peroxidase (GSHPx) activities measured initially and periodically thereafter revealed no difference between Se forms in sheep and horses and only a small (P less than .05) advantage for selenate in supporting serum Se concentration in dairy cattle. Selenium concentrations in skeletal muscle and liver of sheep were not different between Se forms. Serum Se, but not GSHPx, increased with time, and .3 mg of supplemental Se per kilogram of dietary DM from either sodium selenate or sodium selenite supported normal serum Se concentrations in sheep, dairy cattle, and horses.     

RT13型指盘搂草机的设计与研究

针对天然草场产草量低、收集成本高及国外机具价格高这一现状,设计了RT13型指盘搂草机。在介绍搂草机整机结构及工作原理的基础上,对其指盘的运动及指盘平面距离等主要参数进行了设计计算。草场作业试验表明:该搂草机操作简单,形成的草条连续、外形整齐,可有效减少牧草收集时间,从而降低机具燃油消耗及人工成本。该机具各项技术性能指标符合设计要求,满足天然草场的收获要求。     

9BS-2.4苜蓿草种籽播种机的设计研究

针对以紫花苜蓿为代表的豆科牧草种子机械化种植的工艺要求,设计研制了9BS-1.8型苜蓿草种籽播种机。同时,介绍了该机的整体结构、工作原理、技术参数,以及主要部件的设计原理和设计方法。该研究对苜蓿草种籽播种机的研制及相关技术的推广有着指导作用。     

GPRS无线数据传送在农情监测系统中的应用

在精准农业的农情监测系统中,要对各种信息(如GPS地理位置信息、土壤信息等)进行传输。为此,利用GPRS模块Q2406B和单片机W77E58,实现农情监测信息在中国移动GPRS网络上的无线传送。该技术使用方便,成本较低,可以稳定地传送数据,有利于大规模监测网络的建立。     

D型打结器Pro/Engineer仿真及动作功能分析

对D型打结器主要部件的构成及其功能进行了描述,利用Pro/E软件对打结器各元件进行装配,并对其打结装置进行机构的运动模拟仿真.把打结器复杂的打结动作总结为4个打结阶段,最后对打结器主要机构所完成的动作及功能进行了分析,为国内设计制造D型打结器及打结器后续改进进行了前期理论基础工作.     

基于优化的长短时记忆神经网络牧群采食量估测模型

为了能够较为精准的估测牧群的采食量信息,提出一种基于遗传算法(genetic algorithm, GA)和长短时记忆神经网络（long short-term memory,LSTM）的牧群采食量估测模型。首先通过皮尔森系数法分析得出影响牧群的采食量的主要影响因子,以减少输入维度并解决信息冗余问题。在此基础上,构建基于 LSTM 神经网络算法的牧群采食量估测模型,并引入遗传算法来优化LSTM 神经网络模型参数来增加模型的可靠性。最后,利用该模型对牧群采食量进行估测。试验结果表明：该采食量估测模型各评价指标平均绝对误差(mean absolute error,MAE)、平均绝对百分比误差(mean absolute percentage error,MAPE)、以及均方根误差（root mean square error,RMSE）分别为2.982、9.85%和6.108。与单一的LSTM神经网络以及GRU神经网络模型相比,均优于其他模型;且该模型具有较好的估测性能和较强的泛化能力,能够为合理轮牧提供科学指导,对草地保护有一定的应用价值。     

基于地理权重回归模型的土壤有机质空间预测

准确了解土壤有机质的空间分布是合理施肥的重要前提,也是水土流失控制及保护环境的重要基础。利用113个土壤有机质样点数据,以海拔高度、土壤侵蚀强度、土地利用、比值植被指数、样点至河流的欧氏距离、亚铁矿物指数及坡度为参考因子,来尝试利用GWR(Geographically Weighted Regression)模型探索多重因素作用下的土壤有机质空间分布,并通过与普通线性回归(ordinary least squares,OLS)相比较,来了解GWR模型的精度,进而进行了土壤有机质的空间制图,并对其制图效果进行了评价。结果表明,与OLS模型相比,GWR预测模型它能显著降低AIC(Akaike Information Criterion)值,较大程度地提高模型的决定系数,并有效地减少模型的回归残差值。从制图的总体效果看,GWR模型的预测结果与实测值的吻合程度要优于OLS模型。文章还对利用GWR模型进行回归时的样点数量、因子筛选及因子定量化等方面进行了相应的讨论。     

华南兴国县影响土壤侵蚀时空动力学的环境因素

By using soil erosion maps of four different time periods and a digital elevation model (DEM), in combination with the remote sensing and GIS technologies, soil erosion dynamics in Xingguo County of Jiangxi Province in South China were analyzed on both temporal and spatial scales in soils of different parent materials, altitudes and slopes. The results showed that from 1958 to 2000 severe soil erosion was coming under control with a decreasing percentage of the land under severe erosion. It was also found that the soils developed from Quaternary red clay, granite and purple shale were more susceptible to soil erosion and that areas sitting between 200 to 500 m in altitude with a slope less than 3° or between 7° to 20° where human activities were frequent remained to be zones where soil erosion was most likely to occur. These areas deserve special attention in monitoring and controlling.     

基于激光雷达的巡检机器人导航系统研究

智能巡检机器人能够高效、可靠地完成巡检任务,降低工作人员的劳动强度,准确、稳定的导航定位是巡检机器人执行巡检任务的基础。本文研究了基于激光雷达的巡检机器人导航系统,可实现机器人在室内外环境下的地图建立、路径规划和导航定位。导航系统由远程监控平台与巡检机器人组成,远程监控平台发布巡检任务、监控机器人状态、查询与存储检测数据,巡检机器人可实现自主导航定位、遍历检测点、执行数据采集等巡检任务,二者通过无线网络实现远程数据交互。融合激光雷达与编码器信息,使用高鲁棒性Gmapping算法建立二维环境地图。根据地图与检测点信息,采用分支界定算法搜索最优巡检路线,以减少巡检时间和能源消耗。使用自适应蒙特卡罗定位(AMCL)算法估计机器人位置和姿态,结合巡检路线,进行导航定位。根据横向偏差与航向偏差,通过经典的PID算法完成机器人驱动控制。机器人搭载可见光相机与红外相机,可对目标进行可见光通道与红外通道的融合图像检测。对巡检机器人进行了室内导航定位试验,试验结果表明,在1 m/s的速度下,位置与航向偏差的平均绝对误差(MAE)分别小于5 cm和1.1°,标准差(SD)分别小于5 cm和1.5°,能够满足巡检导航定位的要求。