免耕播种机无轴螺旋排肥输肥装置设计与试验

针对西南地区坡度较大、免耕地表秸秆及根茬等造成耕地平整度较差,驱动式破茬防堵免耕播种机作业时机具整体产生振动较大,导致排肥器排肥及导肥管导肥作业性能差的问题,基于螺旋输送原理,设计了一种柔性无轴螺旋排肥输肥装置。通过对肥料的螺旋输送以及物料临界输送速度分析,得出螺旋叶片最佳充肥尺寸以及转速范围。采用EDEM仿真进行二次回归正交旋转试验和响应曲面法分析无轴螺旋排肥输肥装置最佳工作参数：螺旋叶片内半径3mm、螺旋叶片外半径12.8mm、螺旋叶片转速319r/min以及螺旋间距24.5mm。田间测试结果表明,在地表平整度平均值以及地表坡度分别为8.9cm、16.1°时,无轴螺旋排肥输肥装置在作业速度1.5 m/s时,排肥精度误差、均匀性变异系数分别为1.87%、2.52%,满足国家施肥标准,播肥符合当地农艺要求。所设计的无轴螺旋排肥输肥装置满足免耕播种施肥要求,可为在地表平整度较差时排肥和振动较大条件下排肥器以及导肥管的设计与改进提供参考。     

基于改进YOLO v7的生猪群体体温热红外自动检测方法

针对当前生猪规模化养殖过程中基于热红外技术的生猪体温测量效率低的问题,提出了一种基于改进YOLO v7的生猪群体体温检测方法。改进YOLO v7算法在Head层引入VoV-GSCSP结构,降低网络结构复杂度;使用内容感知特征重组(Content-aware reassembly of features,CARAFE)替换模型原始上采样算子,提高特征图放大后的品质,强化生猪头部区域有效特征;引入感受野增强模块（Receptive field enhancement module,RFE）,增强特征金字塔对生猪头部特征的提取能力。本文改进YOLO v7算法对于生猪头部的检测精确率为87.9%,召回率为92.5%,平均精度均值（Mean average precision,mAP）为94.7%。与原始YOLO v7相比,精确率提高3.6个百分点,召回率提高7.0个百分点,mAP提高3.6个百分点。该方法首先自动检测生猪头部区域,再利用头部最大温度与耳根温度的高相关性,最终自动获取生猪体温。温度提取平均绝对误差仅为0.16℃,检测速度为222f/s,实现了生猪群体体温的实时精准检测。综合上述试验结果表明,该方法能够自动定位生猪群体的头部区域,满足生猪群体体温测定的高效和高精度要求,为群养生猪体温自动检测提供了有效的技术支撑。     

豫西传统地坑窑小气候特征及其热舒适性评价

为探究豫西地区地坑窑的热舒适性及影响成因,在冬季和夏季对河南三门峡市曲村的生土地坑窑进行室内外热环境监测,选取相同位置的地上普通砖房建筑作为对照,利用人体热舒适模型对两者室内外热环境和舒适度进行定量分析、对比与评价。结果表明,冬季地坑窑室内空气温度高于地面普通建筑4.16℃,黑球温度高于地面建筑5.17℃,夏季窑室内平均空气温度低于地面建筑3.1℃,黑球温度低4.06℃;冬夏两季地坑窑室内热舒适均优于地上普通建筑,测量期间夏季和冬季窑室内分别有100%和44%的时间处于舒适范围,地面砖房建筑内分别有22%和11%的时间处于舒适范围,其中,夏季热舒适优于冬季;空气温度是影响窑洞室内热舒适的首要因素,黑球温度是影响室外院落热舒适的首要因素;但冬季窑洞内部保温性能较低,夏季通风不畅,室内湿度高达71.97%,冬季窑洞室内温度是更加急需改善的缺陷。     

基于注意力机制的时空卷积数控机床热误差模型研究

为了提高数控机床热误差模型的精度与泛化性,提出了基于注意力机制的长短时记忆卷积神经网络(Long short term memory convolutional neural network based on attention mechanism, AM-CNN-LSTM)热误差模型。利用卷积神经网络提取高维数据空间状态特征的能力和长短时记忆网络提取长时间序列状态特征的能力,构建具有2个支路的热误差模型,分别提取特征后输入到注意力机制中进行特征重要性重构,建立原始数据与热误差的特征映射,最后通过全连接层进行热误差预测。采用G460L型数控机床进行实验数据采集,将不同季节采集到的温度数据和热误差作为模型输入,采用循环学习率与正则化优化方法对模型进行训练。与LSTM、ConvLSTM和CNN-LSTM热误差模型对比,结果表明,AM-CNN-LSTM模型对特征还原能力最强,残差波动范围最小,其残差范围较最大值下降62.09%,模型预测精度在2.4μm以内。     

Sesbania aegyptiaca as promising biomass for manufacturing of MDF

Sesbania fiber is a fast-growing wood species that has been investigated for medium-density fiberboard (MDF) production. To assess the possibility of applying the local industrial defibration parameters of sugar-cane bagasse (SCB) on defibration of sesbania, the chemical constituents of unfibrated and defibrated sesbania, as well as their thermal stability and scanning electron micrographs, were estimated. Different preparation variables of MDF, such as density, level of urea-formaldehyde (UF) resin (with 0.19% free-formaldehyde [HCHO]), and pressing time were studied, in comparison with that produced by using SCB fibers. The results showed that most of the tested sesbania-based MDFs have mechanical properties that fulfill the minimum requirements of MDF ANSI standard. Additionally, applying 12% UF and pressing for 240 sec provided sesbania-based MDF with optimum reduction in thickness swelling (reached ～7%). It is important to note that the sesbania-based MDF produced under these conditions is characterized by a lower TS property, than that obtained from SCB, or that reported in standards. The preliminary feasibility study revealed that using sesbania fibers will be an added economical potential for MDF production.     

不完全分形多孔介质热导率模型

将多孔介质的物理构成分为具有分形结构的团聚体集合和不具有分形特性的固相和孔隙相,建立了简化单元体模型解释其微观结构。结合多孔介质在干燥过程中热量守恒定律和傅里叶导热定律导出了材料总有效热导率模型。此模型无经验常数,每一个参数都有物理意义。研究结果表明,有效热导率与迂曲分形维数、面积分形维数、孔隙率和热风温度呈反比,与热风速率和时间呈正比。     

Effects of temperature and feed composition on essential fatty acid (n-3 and n-6) retention in Atlantic salmon (Salmo salar L.) parr

Retentions of total n-3 and n-6 essential fatty acids (EFAs) were assessed in Atlantic salmon (Salmo salar L.) parr held at 8 °C and 2 °C until they increased in weight from ca. 19 g to 38 g. Feeds contained sandeel oil or a rapeseed:linseed oil blend at 21 and 34% dietary fat. EFA retention efficiencies [(g EFA gained g EFA ingested-1) × 100] were estimated by the 'mass balance method' from measurements of feed intake, changes in biomass for each tank of fish, and fatty acid compositions of the feeds and fish. The n-3 EFA retentions were higher (overall mean 71%) across feed treatments and temperatures than the n-6 EFA retentions (overall mean 63%). Retentions of the n-3 fatty acids were higher in the fish given the feeds with the lower fat content (77% vs. 65%), implying improved retention with reduced n-3 EFA availability. n-3 EFA retention tended to be higher at 2 °C than at 8 °C, although this was not consistent across feeds. At low temperature there was very high retention of the n-3 EFAs in feeds containing sandeel oil (80%). Such high retention may represent an adaptation response to low temperature. Lower n-6 EFA retentions imply that more n-6 fatty acids were metabolized than n-3 EFAs. Feed oil influenced retention of the n-6 fatty acids, retention being lower for the salmon parr given the feeds containing sandeel oil (56% vs. 71%). This could indicate a higher tissue deposition of n-6 fatty acids when they are freely available via the diet. Abbreviations: AA – arachidonic acid (C20:4 n-6); DHA – docosahexaenoic acid (C22:6 n-3); EFA – essential fatty acid; EPA – eicosapentaenoic acid (C20:5 n-3); HUFA – highly-unsaturated fatty acid (\ge4 double bonds); MUFA – monounsaturated fatty acid (1 double bond); PL – phospholipid; PUFA – poly-unsaturated fatty acids (\ge2 double bonds); SFA – saturated fatty acid (no double bond); TAG – triacylglycerol. This revised version was published online in July 2006 with corrections to the Cover Date.