一类具有转点的右端不连续奇摄动边值问题

研究了一类具有转点的右端不连续二阶半线性奇摄动边值问题解的渐近性.首先,在间断处将原问题分为左右两个问题,通过修正左问题退化问题的正则化方程,提高了左问题渐近解的精度,并利用Nagumo定理证明了左问题光滑解的存在性.其次,证明了右问题具有空间对照结构的解,并通过在间断点的光滑缝接,得到了原问题的渐近解.最后,通过一个算例验证了结果的正确性.

三维激光点云数据配准误差传播研究与精度控制

由于测绘科技的飞速发展,三维激光扫描技术在测绘领域的应用日益广泛,但目前还没有通用的三维点云配准误差传播模型,且均未广泛应用到三维激光扫描领域。为此,本文通过引入摄影测量学原理对各测站点云的配准进行分析,探讨了三维激光点云配准误差传播的规律,并将传统控制测量的方法应用到激光扫描测绘中,并优化完善了误差传播模型。通过实验,验证了点云配准过程中配准精度与多站配准误差积累的定量关系,从而对点云配准质量进行评估预测。

一种基于卷积神经网络的VVC去压缩伪影半盲方法

多功能视频编解码(Versatile Video Coding,VVC)是新一代视频编解码标准,拥有较好的压缩性能,能够达到较高的压缩比。但是,编码过程中的变换、量化等操作,不可避免地在视频解码时引起一定程度的压缩伪影,导致解码视频质量降低,影响用户的视觉体验。目前,基于卷积神经网络(Convolutional Neural Network,CNN)的VVC的压缩伪影去除算法并不多,且大部分算法是在默认量化参数已知的情况下建立去伪影模型,对于不知道量化参数的盲场景,这些算法不太适合。直接设计全盲算法是复杂困难的,且性能有限。针对这一情况,提出了一种半盲方法用于去除VVC解码视频中的压缩伪影,该方法比全盲的方法更加灵活且能够达到更好的性能,比非盲方法更加实用。该方法设计出一种分类网络来预测重建视频的量化参数,预训练一些压缩伪影去除模型,根据预测的量化参数为重建视频选择对应的模型以去除压缩伪影。实验结果证明了该算法的有效性。

一类热弹耦合问题的奇摄动解

讨论一类带有小参数的热弹耦合模型,通过奇摄动分析,得到该模型的渐近解。首先,通过相应的边界条件,逐次将外解和边界层矫正项解出,并得出内外解形式的一致有效性;然后,对满足渐近解首项的方程进行数值分析,得出数值求解格式;最后,通过数值算例验证其有效性。

Geometrical design of variable ratio tooth profile based on Boolean subtraction operation and a novel modification method

Variable transmission ratio racks show great potential in rice transplanters as a key component of variable transmission ratio steering to balance steering portability and sensitivity.The objective of this study was to develop a novel geometrical design method to achieve quick,high-quality modeling of the free curvilinear tooth profile of a variable transmission ratio rack.First,a discrete envelope motion 3D model was established between the pinion-sector and the variable transmission ratio rack blank based on the mapping relationship between the rotation angle of the pinion-sector and the displacement of the rack,according to the variable transmission ratio function.Based on the loop Boolean subtraction operation,which removed the pinion-sector from the rack blank during all moments of the discrete motion process,the final complex changing tooth shape of the variable transmission ratio rack was enveloped.Then,since Boolean cutting residues made the variable ratio tooth surface fluctuant and eventually affected the precision of the model,this study proposed a modification method for establishing a smooth and continuous tooth profile.First,a novel fitting algorithm used approximate variable ratio tooth profile points extracted from the Boolean cutting marks and generated a series of variable ratio tooth profiles by utilizing B-spline with different orders.Next,based on a transmission stability simulation,the variable ratio tooth profile with optimal dynamic performance was selected as the final design.Finally,tests contrasting the transmission stability of the machining samples of the initial variable ratio tooth profile and the final variable ratio tooth profile were conducted.The results indicated that the final variable ratio tooth profile is more effective than the initial variable ratio tooth profile.Therefore,the proposed variable ratio tooth profile modeling and modification method for eliminating Boolean cutting residues and improving surface accuracy is proved to be feasible.