分析阶梯流量井流试验数据的线性回归法

文中应用线性回归法，对一次阶梯流量井流试验数据进行适当交换后，可得回归直线方程，与目前采用的多次阶梯流量井流试验数据的分析相比较，在同样可以得到井流实验欲求的井损系数C，综合参数和含水层导水系数T三个参数的情况下，至少可以使抽水量的变化次数减少两次。实际算例表明，一次阶梯流量井流与多次阶梯流量法的计算结果非常吻合。     

多元线性回归和神经网络在水库调度中的应用比较研究

采用多元线性回归和改进人工神经网络算法分别建立水库多目标调度函数，并进行实例计算验证。通过比较分析，表明人工神经网络的非线性映射能力能够更好地反映水库调度中多个自变量和因变量之间的复杂关系，具有较高的模拟精度和较好的可行性，且应用简便。     

A logistic analysis of the factors determining the decision of smallholder farmers to intercrop: A case study involving rubber–tea intercropping in Sri Lanka

The factors influencing the decision of smallholder farmers to adopt new farming technologies were studied with reference to rubber–tea intercropping in Sri Lanka. Rubber–tea intercropping has been recommended previously to rubber farmers as a means to improve productivity and income during the early pre-tapping phase of rubber growth. Although crop trials have shown that the two crops are agronomically compatible and potentially produce a combined economic yield superior to the yield of a sole crop grown on the same area of land, there is little evidence of widespread adoption of this practice among smallholder farmers in Sri Lanka. The aim of the study was to determine the major factors that influence the decision to undertake rubber–tea intercropping and to construct a predictive model that describes the likelihood of adoption of intercropping by traditional smallholder rubber growers. A rapid rural appraisal (RRA) was undertaken based on semi-structured interviews of 90 smallholder farmers in the main rubber growing low wet zone of Sri Lanka. Among a number of factors shown to significantly influence the decision to intercrop tea with rubber, three were shown to operate independently, namely level of income, source of income (i.e. solely from own farm or from farm plus additional off-farm enterprises), and availability of land considered suitable for tea cultivation. A statistical model developed through correlation and logistic analysis, which predicts the likelihood of a smallholder adopting intercropping based on these factors, is presented and discussed. The most likely combination of circumstances (82% probability) under which rubber–tea intercropping is practiced is shown to be where the farmer’s income is greater than Rs. 10,000 per month, where the farmer’s income is based solely on own farm enterprises, and where more than 80% of the farmer’s land area was judged to be suitable for tea cultivation. Conversely, 30% of smallholder farmers that chose not to intercrop did possess land suitable for tea cultivation. Qualitative responses to the RRA indicated that limitation of technical knowledge was the main problem subsequently faced by rubber farmers who had adopted rubber–tea intercropping. Results indicate that there is need for both income support through farm subsidies and further agricultural extension services, if rubber–tea intercropping is to be adopted more widely in Sri Lanka. The wider usefulness of the developed logistic model in determining the likelihood of adoption of intercropping by smallholder farmers is discussed.     

玉溪市林分植被覆盖变化及驱动力分析

林分植被覆盖在很大程度上能够反映林分生态系统的健康与否,其时空变化是林分生态系统发展方向的判断依据。对玉溪市2010—2019年林分植被覆盖的时空格局变化进行了分析,并利用单因素和多元回归分析方法分别建立回归模型探索了林分植被覆盖改善与退化的驱动因素。结果表明:林分植被覆盖整体呈向好趋势,各驱动因素与斜率均不是简单线性关系;道路和居民区与林分植被覆盖改善速度呈负相关、与退化速度呈正相关;DEM、坡度、大型水体、公益林管护与林分植被覆盖改善速度呈正相关、与退化速度呈负相关;坡向与林分植被覆盖改善速度呈正相关,地形起伏度与林分植被覆盖退化速度呈负相关。林分植被覆盖改善与退化的驱动力不尽相同,影响机制也存在差异,这为更好制定生态保护相关政策提供了有力的科学依据。     

稻纵卷叶螟发生程度及发生期预测模型研究

对余庆县1982~2004年稻纵卷叶螟(Cnaphalocis medinalisGuenee)测报资料的筛选分析结果为,第三代幼虫发生程度预测回归模型^y=0.1802 0.0024x1 0.0050x2 0.0185x3±Sy/12=0.6924,第四代幼虫发生程度预测回归模型^y=0.2853 0.0005x1 0.0084x2 0.0112x3±Sy/12=0.6334;第三代低龄幼虫盛发期预测回归模型^y=13.3830 0.6823x±Sy/x=2.61,第四代低龄幼虫盛发期预测回归模型^y=3.1293 0.8776x±Sy/x=2.951。利用其回归模型对稻纵卷叶螟幼虫发生程度和低龄幼虫盛发期进行预测,其方法简便,结果较为准确。     

基于改进YOLOv5算法的农田杂草检测

随着智慧农业技术和大田种植技术的不断发展,自动除草具有广阔的市场前景。关于除草剂自动喷洒的有效性,农田杂草的精准、快速地识别和定位是关键技术之一。基于此提出一种改进的YOLOv5算法实现农田杂草检测,该方法通过改进数据增强方式,提高模型泛化性;通过添加注意力机制,增强主干网络的特征提取能力;通过改进框回归损失函数,提升预测框的准确率。试验表明,在芝麻作物和多种杂草的复杂环境下,本文方法的检测平均精度均值mAP为90.6%,杂草的检测平均精度AP为90.2%,比YOLOv5s模型分别提高4.7%和2%。在本文试验环境下,单张图像检测时间为2.8 ms,可实现实时检测。该研究内容可以为农田智能除草设备提供参考。     

Impact of cash crop cultivation on household income and migration decisions: Evidence from low-income regions in China

This study examines the impact of cash crop cultivation on household income and migration decisions, using survey data collected from low-income regions in China. Given farmers decide themselves whether to cultivate cash crops, an endogenous treatment regression model that accounts for potential selection bias issue is used to analyze the data. The empirical results show that cash crop cultivation exerts a positive and statistically significant impact on household income, but it does not affect household migration decisions significantly. The disaggregated analyses reveal that cash crop cultivation significantly increases farm income but decreases off-farm income.     

土壤因子与小粒咖啡品质产量形成关系研究

土壤因子是影响小粒咖啡品质产量的主要因子。本研究通过对土壤因子进行相关性分析、改进的灰色关联分析、通径分析及多元线性回归分析,明确影响小粒咖啡品质和产量的主要因子,以指导科学施肥。结果表明:小粒咖啡咖啡因含量的土壤决定因子为海拔（-0.296^*）,总糖含量的土壤决定因子为有机质（0.607^**）和速效磷（-0.232^*）,蛋白质的土壤决定因子为pH（-0.387^**）和有机质（0.227^*）,灰分含量的土壤决定因子为速效磷（0.334^**）,水溶性浸出物含量的土壤决定因子为海拔（-0.494^**）,脂肪含量的土壤决定因子为碱解氮（-0.461^**）,产量的土壤决定因子为速效钾（0.339^**）和有机质（0.248^*）。研究结果可为小粒咖啡优质高产的土壤条件筛选和肥料的减施增效提供一定的理论基础,以期增强优质咖啡种质的适宜性和目的性。     

引进甘蔗种质资源杂交后代适应性评价

通过对引进甘蔗种质资源24个杂交后代的的10个农艺性状进行主成分分析、回归分析与通径分析,科学评价引种甘蔗种质资源在广西本地的农艺性状表现,探究各农艺性状间的协同制约关系,为甘蔗种质资源的精准评价和杂交育种材料的科学选配提供理论依据。结果表明:(1)主要农艺性状变异系数为6.79%~29.44%,24个甘蔗种质在分蘖和蔗糖产量上有较大的遗传改良潜力。(2)主成分分析将这10个农艺性状凝聚成4个主成分,分别为产量因子、有效茎因子、糖分因子和出苗因子,贡献率分别是35.2%、16.5%、15.3%、12.1%,累计贡献率达到79.1%;(3)通径分析中各性状对糖产量的直接通径系数大小依次为:蔗茎产量(0.927)>蔗糖分(0.399)>单茎重(0.039)。表明:甘蔗品种(系)GUC23-1、GUC15-2、GUC23、GUC34、GUC25在广西的丰产性最优,在甘蔗引种和选育中,考虑将蔗茎产量、蔗糖分、单茎重作为主要选择性状。