Effect of Additives on Melting Point of LATS Refining Ladle Slag

To avoid slag sticking on the ladle immersion cover during the LATS refining and alloying process, the effect of Al2O3 on the melting point of the ladle slag was studied and the additives including CaF2, B2O3, Li2O, and CaO were used to decrease the melting point of the ladle slag. The melting point was measured using the hemisphere method. The results show that the addition of Al2O3 to the ladle slag increases the melting point. The fluxing action is not remarkable if only CaF2 or CaO is used as the additive. The fluxing action of the composite additive obtained by the mixing of CaO and CaF2 in the mass proportion of ωCaO：ωCaF2=2 ： 1 is preferred. The fluxing action of B2O3 is also notable. When the B2O3 content in mass percent is in the range from 2% to 10%, the corresponding melting point is 1 380 ℃ to 1 290℃. The fluxing action of Li2O is the most remarkable. When the Li2O content is up to 5%, the melting point of the slag is lower than 1 300 ℃.

Application of Rollout Strategy to Test Points Selection for Integer-Coded Fault Wise Table

Test points selection for integer-coded fault wise table is a discrete optimization problem. The global minimum set of test points can only be guaranteed by an exhaustive search which is eompurationally expensive. In this paper, this problem is formulated as a heuristic depth-first graph search problem at first. The graph node expanding method and rules are given. Then, rollout strategies are applied, which can be combined with the heuristic graph search algorithms, in a computationally more efficient manner than the optimal strategies, to obtain solutions superior to those using the greedy heuristic algorithms. The proposed rollout-based test points selection algorithm is illustrated and tested using an analog circuit and a set of simulated integer-coded fault wise tables. Computa- tional results are shown, which suggest that the rollout strategy policies are significantly better than other strategies.

Selection of Minimal Test Points Set for Integer-Coded Fault Wise Table

Test points selection for integer-coded fault wise table is a discrete optimization problem. On one hand, traditional exhaustive search method is computationally expensive. On the other hand, the space complexity of traditional exhaustive is low. A tradeoff method between the high time complexity and low space complexity is proposed. At first, a new fault-pair table is constructed based on the integer-coded fault wise table. The fault-pair table consists of two columns： one column represents fault pair and the other represents test points set that can distinguish the corresponding faults. Then, the rows are arranged in ascending order according to the cardinality of corresponding test points set. Thirdly, test points in the top rows are selected one by one until all fault pair are isolated. During the test points selection process, the rows that contain selected test points are deleted and then the dimension of fault-pair table decreases gradually. The proposed test points selection algorithm is illustrated and tested using an integercoded fault wise table derived from a real analog circuit. Computational results suggest show policies are better than the exhaustive strategy.

基于信息共享策略的加权逐点预测动态多目标优化

为了有效实现Pareto最优解,并且提升对于输入误差的鲁棒性,提出一种基于信息共享策略的加权逐点预测动态多目标优化算法。引入一种信息共享策略,该策略允许每个点利用其相邻解决方案中的信息有效提升模型的鲁棒性。引入一个相似性度量,并对其与一些常用的相似性度量进行对比分析,进一步提出一种加权逐点预测方法,逐点特性极大增强了捕捉各种模式的能力。在EC2018 DMO测试套件上的实验结果验证了该方法的有效性。

基于伪点云特征增强的多模态三维目标检测方法

环境感知是自动驾驶汽车落地的关键技术之一,它对于提高自动驾驶汽车的安全性和可靠性至关重要.三维目标检测是其中的一项核心任务,旨在识别和定位三维空间中的物体,为后续决策提供重要的信息.点云和图像是该任务最常用的输入数据,点云由三维空间中不规则分布的点组成,而图像则是由二维空间上规则分布的像素组成.因此,点云和图像之间难以进行有效的融合.而伪点云作为一种点云表征的图像信息,近几年受到了该领域学者的广泛关注.现阶段基于伪点云的三维目标检测方法还存在伪点云特征提取粗糙和相应感兴趣区域(Region-of-Intersts,RoI)特征表征能力差的问题.本文针对上述问题开展研究,分别提出细粒度注意力卷积和多尺度分组稀疏卷积.细粒度注意力卷积将规则图像处理中常用的深度可分离卷积引入不规则点云的处理流程,并在此基础上嵌入通道和分组注意力机制,进行精细的特征提取,增强伪点云特征;多尺度分组稀疏卷积将格网池化后的Ro I特征分组,进行差异化特征学习,获取不同尺度的Ro I特征,增强伪点云Ro I格网特征的表征能力.基于此,本文在SFD(Sparse Fuse Dense)网络的伪点云特征提取流程中引入细粒度注意力卷积,同时在其伪点云Ro I特征学习流程中引入多尺度分组稀疏卷积,构建SFD++多模态三维目标检测网络.在权威KITTI自动驾驶数据集上的实验结果表明,SFD++每秒可以处理8.33帧数据,其精度在简单、中等和困难的三维汽车检测上达到95.74%、88.80%和86.04%,比次优SFD的精度高出0.15%、0.84%和0.58%.除此之外,一系列消融和补充实验结果验证了所提出卷积的有效性和相关参数设置的合理性.

二维线性化Navier-Stokes-Poisson方程解的逐点估计

本文研究了二维空间线性化的等熵可压缩Navier-Stokes-Poisson方程柯西问题.通过把方程组转变成关于单个函数的方程,求解出各个函数,得到方程组的格林函数.利用对格林函数的详细分析,获得了方程组解的逐点估计.结果显示方程组中电流密度以热核的速度衰减,动量密度衰减慢得多,且其L2范数不衰减.

低碳铝镇静钢精炼渣理化性能研究及动态调控

为提升低碳铝镇静钢水的可浇性,基于低碳铝镇静钢精炼渣的热力学计算,研究典型组分CaO/Al2O3(C/A)、SiO_(2)、FeO对精炼渣液相线温度、1650℃黏度和表面张力的影响规律,结合精炼渣理化性能(熔点和黏度-温度曲线)测试结果,建立吹氩直上精炼渣动态调控模型,并投入生产实践应用。结果表明:随着C/A增加,精炼渣熔点在渣中SiO_(2)质量分数为5%时单调升高,在渣中SiO_(2)质量分数为10%时先降后升,另随C/A的增加精炼渣黏度及黏度-温度曲线的转折温度升高;精炼渣组分范围为C/A 0.75~1.50、SiO_(2)质量分数0~10%、FeO质量分数2%~6%时,可获得较低的液相线温度、较高的黏度和表面张力。生产实践中应用建立的模型可实现造渣物料的动态加入,精炼渣理化性能的稳定性得到显著提升,精炼渣的C/A和全铁(TFe)质量分数标准差明显降低;液相线温度、1650℃时的黏度和表面张力平均值分别降低6.9℃、增加0.0206 Pa·s和0.0109 N/m;全氧(TO)质量分数均值由2.78×10(-3)%下降为2.25×10(-3)%,改进了19.1%,钢水的可浇性得到显著改善。

平滑注意力与谱上采样细化的非等距三维点云模型对应关系计算

为了解决非等距3维点云模型对应关系计算易受模型大尺度形变影响而导致对应失真、准确率低且平滑性差的问题,该文提出一种结合平滑注意力与谱上采样细化的非等距3维点云模型对应关系计算新方法。首先,利用点所在表面的几何特征信息设计平滑注意力机制与平滑感知模块,提高特征对大尺度形变区域非刚性变换的感知能力;其次,将深度函数映射模块与平滑正则化约束相结合,提升函数映射计算结果的平滑性;最后,在谱上采样细化模块中,以多分辨率重建的方式得到最终的逐点映射结果。实验结果表明,与已有算法相比,本算法在FAUST、SCAPE和SMAL数据集上构建的对应关系测地误差最小,处理大尺度形变模型时,能够提升逐点映射的平滑性和全局准确率。

“启智培德”类著作翻译要点探讨

以“启智培德”类著作翻译为研究对象,探讨了中国传统典籍及其现当代衍生作品中“启智培德”类著作翻译三要点——求确义、挖内涵、理文辞。通过对《礼记·大学》和《启迪心灵的钥匙》中的篇章片段进行翻译尝试,探索了对“启智培德”类著作的言外之意进行深入挖掘、对源文本的言表和言外之意进行梳理整合、并在此基础上对“启智培德”类著作翻译三要点予以落实的具体方法,旨在为中国传统典籍类著作外译提供借鉴与参考。

基于距离视图表示与逐点细化结合的点云语义分割方法

三维点云语义分割是机器实现环境感知的重要途径。在现有的研究中,基于体素的算法和基于点的算法在面对大规模的点云数据时计算效率低下。而基于距离视图的算法,在对点云进行投影和反投影时会不可避免地造成精度损失。针对上述问题,提出了基于距离视图表示与逐点细化结合的点云语义分割新框架RPNet。充分的实验表明,所提出的方法在三维点云室外场景数据集SemanticKITTI上的平均交并比达到64.2%,推理速度达到58帧/秒,兼顾了高精度和高速度。

面向3D视觉引导的点云分割

点云分割是三维视觉引导和场景理解中的关键步骤,点云分割的质量直接影响三维测量或成像的质量.为提高分割精度、解决边界越界问题,本文提出了一种面向3D视觉引导的点云分割算法,该算法根据点云的空间位置、曲率和法向量信息,生成初始超体素数据,并提取边界点;通过计算边界点与邻域超体素的相似性度量,进行边界细化,即重新分配边界点优化超体素;最后基于区域生长获得候选片段并根据其凹凸性进行合并,得到对象级分割结果.经过可视化和定量比较表明,该算法有效解决了边界越界问题,能对复杂的点云模型准确分割,分割结果准确率为89.04%,召回率为87.38%.

高牌号无取向硅钢生产中RH精炼渣系熔点分析

通过FactSage软件计算和实验室精炼渣熔点测量高牌号无取向硅钢生产过程中的RH精炼渣熔点,探讨了CaO-SiO_(2)-Al_(2)O_(3)-MgO四元RH精炼渣系中各个化合物对精炼渣熔点的影响。结果显示:当CaO含量为30%~40%时低熔点区域较大;当Al_(2)O_(3)含量为10%~20%时低熔点区域较大;SiO_(2)控制在10%以内较为合适;当MgO含量为8%~10%时低熔点区域较大;误差在5%以下,计算结果可靠。

EAF+VAR生产15-5PH不锈钢点状缺陷成因及改进

采用冶炼成本较低的EAF→VOD→LF+VD→浇注6 t电极棒→真空自耗重熔(VAR)的新工艺流程代替传统双真空冶炼工艺生产15-5PH不锈钢,使用扫描电镜、ASPEX等对其探伤点状缺陷进行检测,并分析其产生原因。结果表明,自耗电极中已存在铝酸钙及镁铝尖晶石非金属夹杂物,真空自耗重熔过程的不稳定,将锭冠和非金属漂浮物卷入熔池残留在自耗锭中,导致锻造成材后产生点状缺陷。通过工艺优化改进,VOD还原渣料由铝粒9 kg/t调整为硅铁8 kg/t、铝粒5 kg/t复合脱氧,精炼渣系成分由(质量分数)CaO为50%~55%、SiO_(2)为10%~15%、Al_(2)O_(3)为20%~25%调整为CaO为45%~50%、SiO_(2)为5%~10%、Al_(2)O_(3)为33%~38%,真空自耗重熔熔速由4.2 kg/min提高至6 kg/min、熔滴时间由0.23 s提高至0.27 s,自耗电极洁净度得到较大的提升,自耗重熔过程熔速控制稳定,熔池液面非金属漂浮物得以去除,成品材高低倍检验均满足标准要求,探伤合格率提高至98%以上。

精细化关键控制点管理在血站血液质量管理中的应用价值

目的探讨血站血液质量管理中精细化关键控制点管理的应用效果。方法选取2022年5月—2023年6月青岛市中心血站采集的680例无偿献血者为研究对象,根据不同管理办法将其分为观察组(n=340)与对照组(n=340)。对照组采用常规的血液质量管理流程。观察组采用精细化关键控制点管理流程,对比2组的血液质量指标,包括血液采集量、血液成分分离率、血液合格率、血液不良反应发生率等。结果观察组血液采集量(200.13±4.82)mL、血液成分分离率[(88.62±2.79)%、(76.52±3.68)%、(94.35±2.23)%]、血液合格率(99.41%)均高于对照组[(197.57±5.35)mL(、85.36±3.24)%、(72.37±4.11)%、(91.12±2.56)%、97.35%],差异有统计学意义(P<0.05)。观察组的血液不良反应发生率(0.59%)低于对照组(2.94%)(P=0.020)。结论实施精细化关键控制点管理可以有效提升血液质量,降低不良反应,对保障血液安全和供应具有积极意义。

自制复合细化剂对废杂黄铜制备铜锭组织性能的影响

主要研究自主开发的JT-1复合细化剂对于废杂黄铜熔炼组织性能的影响,重点分析废杂黄铜熔炼过程的JT-1复合细化剂在解决细化中毒问题的机理,同时探究了JT-1复合细化剂使用量的变化对于晶粒细化、消除枝晶、均匀化组织以及改善熔体流动性的影响。研究结果表明,JT-1复合细化剂加入质量分数在0.03%~0.04%,可以构建多重细化环境,废杂黄铜熔炼熔体凝固结晶晶粒度由100μm以上降低到35μm,同时熔体流动性也得到明显改善,铸锭抛光硬质点也在2个以内达到一级以上的标准,完成满足废杂黄铜生产环保卫浴铜锭产品质量要求。

A Fuzzy Decision Based WSN Localization Algorithm for Wise Healthcare

Wise healthcare is a typical application of wireless sensor network(WSN), which uses sensors to monitor the physiological state of nursing targets and locate their position in case of an emergency situation. The location of targets need to be determined and reported to the control center,and this leads to the localization problem. While localization in healthcare field demands high accuracy and regional adaptability, the information processing mechanism of human thinking has been introduced,which includes knowledge accumulation, knowledge fusion and knowledge expansion. Furthermore, a fuzzy decision based localization approach is proposed. Received signal strength(RSS) at references points are obtained and processed as position relationship indicators, using fuzzy set theory in the knowledge accumulation stage; after that, optimize degree of membership corresponding to each anchor nodes in different environments during knowledge fusion; the matching degree of reference points is further calculated and sorted in decision-making, and the coordinates of several points with the highest matching degree are utilized to estimate the location of unknown nodes while knowledge expansion. Simulation results show that the proposed algorithm get better accuracy performance compared to several traditional algorithms under different typical occasions.

Global Fine Registration of Point Cloud in LiDAR SLAM Based on Pose Graph

The laser scanning system based on Simultaneous Localization and Mapping(SLAM)technology has the advantages of low cost,high precision and high efficiency.It has drawn wide attention in the field of surveying and mapping in recent years.Although real-time data acquisition can be achieved using SLAM technology,the precision of the data can’t be ensured,and inconsistency exists in the acquired point cloud.In order to improve the precision of the point cloud obtained by this kind of system,this paper presents a hierarchical point cloud global optimization algorithm.Firstly,the“point-to-plane”iterative closest point(ICP)algorithm is used to match the overlapping point clouds to form constraints between the trajectories of the scanning system.Then a pose graph is constructed to optimize the trajectory.Finally,the optimized trajectory is used to refine the point cloud.The computational efficiency is improved by decomposing the optimization process into two levels,i.e.local level and global level.The experimental results show that the RMSE of the distance between the corresponding points in overlapping areas is reduced by about 50%after optimization,and the internal inconsistency is effectively eliminated.

LATS refining ladle slag modifying with CaO-CaF_2

To reduce the slag sticking onto the snorkel of the ladle during the ladle alloying treatment station （LATS） process, CaO- CaF2 （the mass ratio of CaO/CaF2 is 1：1） was employed as the modifier of the LATS refining ladle slag. The effect of CaO-CaF2 on the melting point, viscosity, and desulfurizing capability of the ladle slag was investigated. The melting point of the unmodified ladle slag is 1439℃. When adding 20wt% CaO-CaF2, the melting point is decreased to 1327℃. At 1500℃. the viscosity of the unmodified ladle slag is 6.5 Pa.s, which can be decreased lower than 2 Pa.s by adding more than 10wt% CaO-CaF2. The experimental results of desulfu- rization of the melts show that the desulfurizing power of the ladle slag can be enhanced by adding CaO-CaF2.

Better Refined Adsorption Isotherm than BET Equation

During studying the heat capacity of metals and brightening more than the original Lena’s image, the temperature increasing term obtained in binomial expansion is transformed into the adsorption increasing term and thereafter we have derived the total adsorption rate equation with it. In the first layer the quantization does not occur and from 2nd layer to nth layer the quantization occurs. So as to get the total adsorption rate equation we add the quantized terms of the second to nth layers to the non-quantized term of the first layer. All terms are based on the unit surface sites. Instead of the unit surface sites, the new adsorption site term appears in the denominator of the adsorption equation. Hence the adsorption equations come out much better than BET equation. The surface area is also calculated through the integration of the adsorption isotherm equation excluding the first layer adsorption equation from the inflection point to the wanted relative pressure.

基于边界极限点特征的改进YOLOv3目标检测

目标数量多、尺度较小与高度重叠等问题导致目标检测精度低、难度大。为提升目标检测精度,尽可能避免漏检、误检情况,提出一种基于边界极限点特征的改进YOLOv3目标检测算法。首先,引入边界增强算子Border,从边界的极限点中自适应地提取边界特征来增强已有点特征,提高目标定位准确度;然后,增加目标检测尺度,细化特征图,增强特征图深、浅层语义信息的融合,提高目标检测精度;最后,基于目标检测中目标实例特性及改进网络模型,引入完全交并比(CIoU)函数对原YOLOv3损失函数进行改进,提高检测框收敛速度以及检测框召回率。实验结果表明,相较于原YOLOv3目标检测算法,改进后的YOLOv3目标检测算法的平均精度提高了3.9个百分点,且检测速度与原算法相近,能有效提高模型对目标的检测能力。