基于ANSYS的某自卸车车架有限元分析

为了辅助某自卸车车架的设计,对此车架进行了模型的简化与建立。依靠ANSYS对装配后的车架进行了全局与局部的网格划分;在此基础上对弯曲工况、扭转工况、制动工况以及过减速带工况这四种典型工况进行了车架应力与变形分析。通过对变形位移图与应力分布图的分析得出了车架容易出现失效的区域,从而为车架强度设计提供了依据。     

汽车侧碰模型简化及其材料优化研究

根据整车侧碰模型建立了简化模型,通过B柱上、中、下部的变形速度和B柱的变形模式对比验证了简化模型的有效性。通过对简化模型的多因素正交优化设计,总结了侧面结构材料参数对侧碰性能的影响趋势,依次为:B柱材料、门槛件厚度、B柱加强件材料、门槛件材料、B柱加强件厚度,并将优组合方案应用于整车模型,仿真结果表明了材料优化方案的B柱变形速度趋势和原模型保持一致,假人胸部上肋骨峰值保持不变,中肋骨峰值得到了降低,下肋骨峰值有所升高,但在法规允许的范围内。     

拖拉机冷却系统效率影响因素分析

研究了拖拉机冷却系统效率的影响因素,为提高整机冷却系统效率提供技术支持。     

并联式混合动力汽车瞬时控制策略研究与改进

通过对并联式混合动力汽车瞬时控制策略的研究,分别建立了动力电池在充电和放电两种工况下的等效油耗函数,考虑到动力电池SOC和混合动力汽车制动能量回收这两个因素的影响,导出了改进后的等效油耗瞬时控制目标函数,并建立了瞬时控制模型。针对某款并联式混合动力汽车,运用ADVISOR 2002软件,对改进前后的瞬时控制策略进行了仿真。     

红枣气体射流冲击干燥收缩特性研究

Chinese jujube originated in China and has a long historical standing. China is the only country which can produce a great amount of Chinese jujube in the world. Modern research showed that the nutrition of Chinese jujube is abundant. Chinese jujube has great medicinal and food value, with high economic value and good prospects for development. Shrinkage characteristics of dates generally refer to volume shrinkage characteristics of whole dates. Large fruit and plump grain appearance of dates is the first choice of consumers to buy elements. The volume of dried jujube points to the consumer’s heart, and is an important indicator of consumers whether to buy. Jujube processing enterprises grade the date in accordance with its size, different grades dates have different value and price. Therefore, drying is the most important steps during the dates processing as it affects directly the quality of the product, and the shrinkage characteristics is the most important quality characteristic of Chinese jujube after drying. However, the traditional natural desiccation and desiccation in the drying room of the red dates has many problems. The conventional drying method for Chinese jujube has some disadvantages of severe shrinkage, long drying time and large energy consumption etc. Therefore, in order to improve the quality of dried Chinese jujube and reduce the shrinkage, it is imperative to replace the traditional drying method with the modern technologies. Air impingement drying technology was applied to dry Chinese jujube. The drying characteristics of shrinkage, moisture effective diffusivity considering shrinkage characteristics and the activation energy of Chinese jujube shrinkage were investigated under different air temperatures (55, 60, 65 and 70℃) and air velocities (6, 9, 12 and 15m/s). The dynamic changes of the regular pattern for the shrinkage characteristic of Chinese jujube in the air impingement drying process were studied, the results showed that the volume relative shrinkage ratio of Chinese jujube appeared to increase and then decrease with the increase of air temperature and air velocity. Considering shrinkage characteristics, the moisture effective diffusivity of Chinese jujube in the air impingement drying process through the Fick’s second law was obtained. The diffusivity appeared to increase and then decrease with the decrease in moisture content in dry basis of Chinese jujube. The Arrhenius equation calculated the shrinkage activation energy of the Chinese jujube was 1281kJ/mol. This research provided technical basis for shrinkage characteristics of Chinese jujube using air impingement drying technology.     

基于Lab颜色空间的非监督GMM水稻无人机图像分割

为解决传统水稻冠层图像分割算法性能在很大程度上依赖于训练数据集的质量,且分割效果易受田间多变光照强度影响,导致水稻生产参数估计精度不高等问题,提出一种基于Lab颜色空间的非监督贝叶斯方法,用于田间水稻无人机图像分割.模型参数从每个独立、未标记的无人机图像直接学习获得,无需训练.不同图像会有不同的模型参数,该算法能够适应...     

基于无人机多时相植被指数的冬小麦产量估测

通过无人机搭载多光谱相机,对不同水分亏缺条件下冬小麦多个生育期进行遥感监测,采用不同种类多光谱植被指数表征冬小麦的生长特征,分析了植被指数与冬小麦产量的相关关系,并利用多时相植被指数构建产量估测数据集,采用偏最小二乘回归、支持向量机回归和随机森林回归3种机器学习算法进行冬小麦产量估测.结果表明,随着冬小麦的生长,多个植...     

基于无人机遥感与随机森林的荒漠草原植被分类方法

荒漠草原是草原中最旱生的类型,属于草原的极限生态状态,也是气候变化和生态系统演变的预警区。利用无人机高光谱遥感技术快速、准确地提取荒漠草原草地植被类型,对动态监测草原生态安全和合理开发草地畜牧业具有重要意义。以无人机搭载高光谱成像系统采集内蒙古荒漠草原遥感图像,获得具有高空间分辨率和高光谱分辨率的图像;通过光谱连续统去除变换,增强草地植被之间的光谱差异,并构建植被指数;采用分步波段选择法选择荒漠草原植被的特征波段,实现高光谱数据降维;构建融合光谱特征、植被特征、地形特征和纹理特征等24个变量的随机森林分类模型,并与支持向量机（SVM）、K-最近邻（KNN）和最大似然分类（MLC）法进行比较。结果表明,在4种分类方法中随机森林分类算法分类效果最好,总体分类精度达到91.06%,比SVM、KNN和MLC等机器学习算法分别高7.9、15.61、18.33个百分点,Kappa系数达到0.90,比SVM、KNN和MLC算法分别高0.13、0.23和0.26。无人机高光谱低空遥感和随机森林算法的结合为荒漠草原草地植被分类提供了新途径。     

鸡体鸡笼阻挡对密闭鸡舍气流运动影响的数值研究

在建立密闭式鸡舍机械通风系统气流运动方程的基础上,推导了由流函数和势函数表示的气流运动控制方程。根据能量变分原理,导出了平面等参可变节点的统一有限元模型。重点讨论了流函数形式下的多连通域(即鸡舍内存在阻挡物)方程的数值分析法。提出了采用流场叠加法求解时的两种内边界条件确定方法:①切向循环速度为零条件:在任一封闭内边界上切向循环速度为零;②能量最小条件:真实内边界条件使整体流场的能量最小。通过上述两种条件建立了原方程的补充代数方程,使问题得到了圆满解决。运用数值模拟,比较了横向通风与纵向通风系统在舍内气流分布型式上的特点。结果表明,在纵向通风形式下,鸡舍内的气流分布非常均匀,鸡体鸡笼等障碍物对空气运动的阻挡作用不十分明显。相反,横向通风系统的气流分布极不均匀,尤其下部鸡笼周围的气流速度大大降低。     

基于增强Tiny YOLOV3算法的车辆实时检测与跟踪

针对深度学习方法在视觉车辆检测过程中对小目标车辆漏检率高和难以实现嵌入式实时检测的问题,该文基于Tiny YOLOV3算法提出了增强Tiny YOLOV3模型,并通过匈牙利匹配和卡尔曼滤波算法实现目标车辆的跟踪。在车载Jetson TX2嵌入式平台上,分别在白天和夜间驾驶环境下进行了对比试验。试验结果表明:与Tiny YOLOV3模型相比,增强Tiny YOLOV3模型的车辆检测平均准确率提高4.6%,平均误检率减少0.5%,平均漏检率降低7.4%,算法平均耗时增加43.8 ms/帧;加入跟踪算法后,本文算法模型的车辆检测平均准确率提高10.6%,平均误检率减少1.2%,平均漏检率降低23.6%,平均运算速度提高5倍左右,可达30帧/s。结果表明,所提出的算法能够实时准确检测出目标车辆,为卷积神经网络模型的嵌入式工程应用提供了参考。