Bacterial community structure of nirK-bearing denitrifiers and the development of properties of soils in created mitigation wetlands

We investigated the abundance and genetic heterogeneity of bacterial nitrite reductase genes (nir) and soil structural properties in created and natural freshwater wetlands in the Virginia piedmont. Soil attributes included soil organic matter (SOM), total organic carbon (TOC), total nitrogen (TN), pH, gravimetric soil moisture (GSM), and bulk density (D_b). A subset of soil attributes were analyzed across the sites, using euclidean cluster analysis, resulting in three soil condition (SC) groups of increasing wetland soil development (i.e., SC1 < SC2 < SC3; less to more developed or matured) as measured by accumulation of TOC, TN, the increase of GSM, and the decrease of D_b. There were no difference found in the bacterial community diversity between the groups (p = 0.4). NirK gene copies detected ranged between 3.6 × 10⁴ and 3.4 × 10⁷ copies g⁻¹ soil and were significantly higher in the most developed soil group, SC3, than in the least developed soil group, SC1. However, the gene copies were lowest in SC2 that had a significantly higher soil pH (~6.6) than the other two SC groups (~5.3). The same pattern was found in denitrifying enzyme activity (DEA) on a companion study where DEA was found negatively correlated with soil pH. Gene fragments were amplified and products were screened by terminal restriction fragment length polymorphism (T-RFLP) analysis. Among 146 different T-RFs identified, fourteen were dominant and together made up more than 65% of all detected fragments. While SC groups did not relate to whole nirK communities, most soil properties that identified SC groups did significantly correlate to dominant members of the community.     

基于BP神经网络的柴油发动机特性建模

建立连续的发动机燃油特性和调速特性数学模型作为液压机械无级变速器虚拟试验平台的动力源。根据虚拟平台对不同特性区域的精度需求对柴油发动机不同特性区域的试验数据进行不同的密度选取、乱序和归一化处理,采用单隐层BP神经网络对试验数据进行训练,对比不同隐层节点数网络的训练误差和测试误差,选取误差最小的网络,求解出网络的数学表达式。通过该方法以ISLe310柴油发动机为例建立燃油特性和调速特性的连续数学模型,这两个简单的数学表达式准确反映了发动机万有特性和外特性,连续模型避免了虚拟试验中出现信号的突变和奇异点。通过和经典的最小二乘法拟合得到的最优特性模型进行对比,其具有更小的误差、更强的泛化能力,能够更好地反映柴油发动机的相关特性。     

液压机械无级变速传动拖拉机虚拟试验平台设计

采用计算虚拟试验技术搭建无级变速拖拉机虚拟试验平台,缓解液压机械无级变速器(HMCVT)研究中实物试验存在的成本高、周期长及试验场地限制问题.以模块化和参数化建模的方法,将拖拉机系统划分为发动机模块、HMCVT模块、中央传动模块及行走机构模块,搭建了虚拟试验平台.采用神经网络建立发动机模型,采用动力学方程建立液压机械无...     

基于改进区域卷积神经网络的田间玉米叶部病害识别

目的 引入区域卷积神经网络Faster R-CNN算法并对其改进，以实现在田间真实环境下背景复杂且具有相似病斑特征的玉米病害的智能诊断。方法 在玉米田间和公开数据集网站获取具有复杂背景的9种常见病害图像1 150幅，人工标注后对原始图像进行离线数据增强扩充；对Faster R-CNN算法进行适应性改进，在卷积层加入批标准化处理层，引入中心代价函数构建混合代价函数，提高相似病斑的识别精度；采用随机梯度下降算法优化训练模型，分别选取4种预训练的卷积结构作为Faster R-CNN的特征提取网络进行训练，并测试得到最优特征提取网络，利用训练好的模型选取不同天气条件下的测试集进行对比，并将改进Faster R-CNN与未改进的Faster R-CNN和SSD算法进行对比试验。结果 在改进Faster R-CNN病害识别框架中，以VGG16卷积层结构作为特征提取网络具有更出色的性能，利用测试集图像检验模型，识别结果的平均精度为 0.971 8，平均召回率为0.971 9，F1为0.971 8，总体平均准确率可达97.23%；晴天的图像识别效果优于阴天的。改进Faster R-CNN算法与未改进的Faster R-CNN算法相比，平均精度高0.088 6，单张图像检测耗时减少0.139 s；与SSD算法相比，平均精度高0.0425，单张图像检测耗时减少0.018 s，表明在大田环境中具有复杂背景的玉米病害智能检测领域，改进Faster R-CNN算法综合性能优于未改进的Faster R-CNN算法和SSD算法。结论 将改进后的Faster R-CNN算法引入田间复杂条件下的玉米病害智能诊断是可行的，具有较高的准确率和较快的检测速度，能够避免传统人工识别的主观性，该方法为田间玉米病害的及时精准防控提供了依据。     

ET_0预测的卡尔曼滤波修正ANFIS模型研究

实时、准确地对作物需水量的预测是实现智能节水灌溉的关键技术。预测模型的合理选择及精度提高是作物需水决策系统的核心。本文将陕西西安地区的气象数据环境信息引入自适应神经模糊推理(ANFIS)作物参考蒸腾量(ET_0)预测模型,应用卡尔曼滤波器对气象数据经ANFIS建模得到的ET_0预测值进行滤波去噪,以提高模型的预测精度,并通过仿真和实验验证,从理论和实践两个方面来验证模型的精度。仿真结果得到,反映模型预测值与真实值之间拟合程度的均等系数(EC)值校正前为0.93,校正后达到0.98。实验结果得到,ANFIS预测模型的平均绝对误差是28.94%,平均相对误差是26.37%,卡尔曼修正后的ANFIS预测模型的平均绝对误差是7.24%,平均相对误差是6.59%。仿真和实验结果表明,利用卡尔曼滤波对ANFIS预测模型进行修正,可以提高预测的精度,经卡尔曼修正后的ANFIS模型能更佳地反映ET_0的变化趋势。     

提高燃油经济性的拖拉机变速控制策略

为实现对液压机械无级变速拖拉机的最佳燃油经济性控制,分析发动机及液压机械无级变速器(hydro-mechanical continuously variable transmission, HMCVT)对拖拉机燃油经济性的影响,研究拖拉机最佳燃油经济性无级变速控制策略,该文针对发动机和HMCVT二元调节无级变速拖拉机,分析了发动机燃油消耗率和变速器的效率变化特性,提出了以发动机有效燃油消耗率g_e与HMCVT传动效率η_b的比值g_e/η_b为指标的最佳燃油经济性无级变速控制策略及拖拉机负载反馈控制原理。采用参数循环算法,求算出拖拉机在负载特性场内任意工作点下的最佳发动机转速、转矩、HMCVT的最佳变速比,保证了二元协同调节下拖拉机最佳燃油经济性变速控制策略的工程实现。计算结果显示:最佳变速比的分布呈现梯田状,平台部分的最佳变速比对应HMCVT纯机械传动时的工作状态,此时HMCVT处于传动效率最高点,并且在变速器传动效率高于0.92的工作区,最佳变速比的分布比例高达72.84%。相比较一元调节下分别以g_e、g_e/η_b为指标、二元调节下以g_e为指标的3种变速控制策略,明显降低了拖拉机燃油消耗率。牵引功率范围内,当拖拉机在某一目标车速下稳定工作时,在基于g_e/η_b最小化的二元调节变速控制策略调控下,拖拉机更可能在较低油耗状态下工作。表明以g_e/η_b为指标的二元调节拖拉机最佳燃油经济性变速控制策略能够提高拖拉机在任意工况下的燃油经济性。     

拖拉机机电复合无级变速器传动特性研究

提出一种拖拉机用机电复合无级变速器方案(Electromechanical Compound Continuously Variable Transmis-sion,EMCVT),将发动机动力分流为两路,机械路高效传动大功率,电机路实现无级变速,二路通过双行星排汇流完成高效无级变速传动.分析了传动系统的速比特性、转矩特...     

Structural optimization of labyrinth-channel emitters based on hydraulic and anti-clogging performances

To balance the hydraulic and anti-clogging performances for labyrinth-channel emitters, a multi-objective optimization was performed to improve the structure of trapezoidal labyrinth channels. Two variables, pressure loss coefficient (PLC) and passage rate of particles (PRP), were proposed to evaluate the hydraulic and anti-clogging performances, respectively. The Pareto fronts of PLC and PRP were calculated using the elitist non-dominated sorting genetic algorithm version II (NSGA-II). The results show that the optimal fronts are obtained through 100 generations evolution. The associated key parameters D/W and H/D are constant, while S/H and α are different for each front. In addition, the approach to obtain a trade-off solution from these fronts was developed by arranging different weights for the two variables. This approach is very suitable for structural optimization of labyrinth channels with the same shape but different dimensions.