成熟期梨可溶性固形物含量的近红外漫反射光谱无损检测

【目的】研究成熟期梨可溶性固形物含量的近红外漫反射光谱无损检测技术,为及时、准确地掌握成熟期梨果实的内部品质特性及田间管理、适时采收、合理储藏提供依据。【方法】基于近红外漫反射光谱检测技术分别建立了成熟期砀山酥梨可溶性固形物含量的偏最小二乘(PLS)、广义回归神经网络(GRNN)和偏最小二乘支持向量机动态预测模型(LSSVM),并综合评价了无信息变量消除法(UVE)优选有效特征波数对于简化模型、提高预测性能的影响。【结果】UVE算法能够很好地提高建模效率、有效改善GRNN和LSSVM模型预测精度,而对PLS分析模型效果不明显。3种模型中,LSSVM模型比GRNN和PLS模型具有明显优势,其中UVE-LSSVM模型具有最佳预测精度和适用性,其校正相关系数(Rc)为0.988,校正均方根误差(RMSEC)为0.074,预测相关系数(Rp)为0.922,预测均方根误差(RMSEP)为0.162。【结论】基于近红外光谱技术的UVE-LSSVM模型可用于成熟期梨可溶性固形物含量的无损检测。     

Random within-herd variation in financial performance and time to financial steady-state following management changes in dairy herd

Abstract

The manager of a dairy herd and the affiliated consultants constantly need to judge whether financial performance of the production system is satisfactory and whether financial performance relates to real (systematic) effects of changes in management. This is no easy task because the dairy herd is a very complex system. Thus, it is difficult to obtain empirical data that allows a valid estimation of the random (within-herd) variation in financial performance corrected for management changes. Thus, simulation seems to be the only option. This study suggests that much caution must be recommended when claming effect of changes in herd management because the link between management changes (cause) and effect (measured as improvement of gross margin per cow year) is extensively blurred by a large within-herd variation in available real life accounting data and differences between herds in time to steady state following management changes.     

Effect of pressure support ventilation during weaning on ventilation and oxygenation indices in healthy horses recovering from general anesthesia

ObjectiveTo determine if pressure support ventilation (PSV) weaning from general anesthesia affects ventilation or oxygenation in horses.Study designProspective randomized clinical study.AnimalsTwenty client‐owned healthy horses aged 5 ± 2 years, weighing 456 ± 90 kg.MethodsIn the control group (CG; n = 10) weaning was performed by a gradual decrease in respiratory rate (f_R) and in the PSV group (PSVG; n = 10) by a gradual decrease in f_R with PSV. The effect of weaning was considered suboptimal if PaCO₂ > 50 mmHg, arterial pH < 7.35 plus PaCO₂ > 50 mmHg or PaO₂ < 60 mmHg were observed at any time after disconnection from the ventilator until 30 minutes after the horse stood. Threshold values for each index were established and the predictive power of these values was tested.ResultsPressure support ventilation group (PSVG) had (mean ± SD) pH 7.36 ± 0.02 and PaCO₂ 41 ± 3 mmHg at weaning and the average lowest PaO₂ 69 ± 6 mmHg was observed 15 minutes post weaning. The CG had pH 7.32 ± 0.02 and PaCO₂ 57 ± 6 mmHg at weaning and the average lowest PaO₂ 48 ± 5 mmHg at 15 minutes post weaning. No accuracy in predicting weaning effect was observed for f_R (p = 0.3474), minute volume (p = 0.1153), SaO₂ (p = 0.1737) and PaO₂/PAO₂ (p = 0.1529). A high accuracy in predicting an optimal effect of weaning was observed for V_T > 10 L (p = 0.0001), f_R/V_T ratio ≤ 0.60 breaths minute^?1 L^?1 (p = 0.0001), V_T/bodyweight > 18.5 mL kg^?1 (p = 0.0001) and PaO₂/FiO₂ > 298 (p = 0.0002) at weaning. A high accuracy in predicting a suboptimal effect of weaning was observed for V_T < 10 L (p = 0.0001), f_R/V_T ratio ≥ 0.60 breaths minute^?1 L^?1 (p = 0.0001) and Pe′CO₂ ≥ 38 mmHg (p = 0.0001) at weaning.Conclusions and clinical relevancePressure support ventilation (PSV) weaning had a better respiratory outcome. A higher V_T, V_T/body weight, PaO₂/FiO₂ ratio and a lower f_R/V_T ratio and Pe′CO₂ were accurate in predicting the effect of weaning in healthy horses recovering from general anesthesia.     

基于DEA模型的广东财政支农资金使用效率分析

以广东省为研究对象,研究了广东省财政支农资金的使用效率。运用主成分分析法（PCA）对数据进行处理,取消其变量间的相关性,保证运用DEA 模型测算使用效率结果的准确性。应用DEA 方法,以VRS 模型计算效率值,得到技术效率、纯技术效率和规模效率。分析过程主要从两个角度进行：一是对2011 年广东省除深圳外的20 个城市财政支农资金的使用效率进行横向分析,结果表明广东省大部分城市财政支农效率较低,存在改进的空间。二是对1995 年税制改革以来广东全省的财政支农资金使用效率进行纵向分析,结果表明自1995 年以来,广东省财政支农资金使用效率总体呈下降趋势,2010 年以后有所回升。     

基于支持向量机的水稻稻瘟病图像分割研究

水稻稻瘟病图像的分割是水稻稻瘟病自动分析与识别的关键环节,其分割效果直接影响后续处理。提出一种基于支持向量机的水稻稻瘟病病害彩色图像分割方法。首先选取叶子正常部分的像素点以及颜色相对复杂的病斑像素点作为负训练样本和正训练样本,提取像素R、G、B彩色分量作为特征向量,对支持向量机进行训练,然后在RGB空间利用训练好的支持向量机对待分割图像的所有像素点进行分类,实现水稻稻瘟病彩色图像的分割。为了获得最佳的分割效果,采用网格搜索法对径向基核函数下的不同核参数分割效果和性能进行比较与分析,确定最佳模型参数。利用此模型进行水稻稻瘟病图像分割实验,获得较好的分割精度,结果优于最大类间方差分割算法。     

天津市都市农业发展对财政金融支持需求的调研分析

都市农业的发展离不开财政与金融的强力支持.调研了天津市都市农业发展对财政金融支持的需求特征,结果表明,农户和农业园区对农业技术的需求是多样的,需要政府财政与金融对农业发展的大力支持,财政和金融合力支持农业发展是今后支农机制创新的方向.     

基于颜色和纹理特征的大麦主要病害识别研究

为提出一种在自然环境条件下基于采集图像的颜色一阶与二阶矩和纹理LBP算子改进模式综合特征参数的大麦病害识别方法,以甘肃河西地区发生的大麦白粉病、云纹病和条锈病为研究对象,采用颜色矩和LBP算子均匀模式综合特征参数来提取大麦病斑的颜色和纹理特征,并将该特征向量作为输入向量构建以径向基为核函数的支持向量机(SVM)分类器模型。利用SVM分类模型对采集到的355幅病害图像进行实例分析,结果表明当径向基参数时,大麦病害整体识别正确率达84.7458%。本研究为农田大麦病害诊断提供了有效的分析手段,验证分类模型在大麦病害研究中的可行性,并可为其他农作物病害诊断提供借鉴和参考。     

Deformation Analysis on the Cuplock of High Support System

The Bowl buckle high support system is widely applied to the actual construction with its convenient construction and high bearing capacity.At present, because of the lack of erection norms and the personnel experience in construction of high support system, so the erection of the structural measures are not reasonable, the system deformation under the action of load is too large which would impact the construction quality.Using ANSYS finite element software simulate the effects of the high support system deforation which arranged in different direction scissors and fixed to the surrounding structure.The vertical and horizontal scissors are arranged to system of vertical and lateral displacement with high support effective constraint.The deformation of the high support system which fixed to the surrounding structure reduced, displacement boundary portion of the vertical rod decreased more obviously.     

Adaptive Inverse Control for Feed Water and Superheated Steam Temperature of Supercritical Pressure Boiler

An adaptive inverse control method for feed water and superheated steam temperature of supercritical pressure boiler is investigated. Through the analysis of the boiler system ,s dynamic characteristics, the input vector structure of inverse dynamics model of feed water and superheated steam temperature is constructed. The adaptive inverse control system is realized through on-line identification of inverse dynamic model with recursive least squares support vector machine. In the control process, the demand for feed water and spray water flow to the outlet steam temperature and micro-superheated steam temperature is considered to realize the integrated control of feed water and superheated steam temperature. Simulation results show that this control system has good control performance and adaptability and can void the repeated oscillation phenomena of control variables which appear in present two section control system.