Enhanced Thermal Conductivity of 5A Molecular Sieve with BNs Segregated Structures

5A molecular sieves have been widely used as adsorbents in cryogenic distillation for hydrogen isotope separation in fusion reactor engineering, but its low thermal conductivity is detrimental to the process stability. Improving the thermal conductivity of 5A molecular sieves is one of the most important goals for high‐performance devices. Here, firm segregated structures with boron nitride sheets (BNs) are constructed around 5A molecular sieve particles. SEM results show 30 µm BNs tend to form the better networks in comparison with that of 0.12 µm BNs at 40 wt% loadings. It is further verified that BNs with the larger size promote the thermal conductivity. Meanwhile, the thermal conductivity increases with the increasing amounts of BNs. XRD and specific surface area results indicate that the sintering and the addition of BNs exert negligible effects on the structure of 5A molecular sieve. These results indirectly show 5A molecular sieve with BNs segregated structures is very likely to be used for the application of hydrogen isotopic separation. Besides, this work provides new insight into the construction of segregated structure in inorganic porous materials.

     

Effect of nozzle type on pulse-jet cleaning performance of ceramic filter tube

The dust-cleaning efficiency of the ceramic filter tube is the key factor to determine the long-term stable operation of the dust removal system, and the type of nozzle significantly influences the performance. In this study, four types of nozzles (injection hole, sudden enlargement nozzle, tubular nozzle, and induction nozzle) are used to conduct comparative experimental study on the dust-cleaning effect of the ceramic filter tube. Using the schlieren technique find that the tubular and induction nozzles can improve the deviation angle of pulse-jet airflow. The injection parameters were set as an injection pressure of 200 kPa, an injection distance of 120 mm and a pulse width of 80 ms in the experiment. Using the sidewall pressure measurement system indicates that the integrated pressure of the sudden enlargement nozzle is the largest, but the uniformity is poor. The comprehensive pressure of the sudden enlargement nozzle is increased by 64.86 %, and the non-uniformity is reduced by 28.62 % compared with those of the induction nozzle. Using the visualization filtration cycle test system implies that the degree of damage and stripping state of the dust layer are directly proportional to the sidewall pressure. Moreover, it is found that the residual pressure drop changes exponentially with the dust removal quality under different nozzle types. These conclusions can help explain the mechanism of dust stripping and provide a basis for the selection of nozzles.     

基于梯度遗传算法的压电换能器稳健优化设计

悬臂梁式压电振动换能器是一种新型的供电装置,其设计受到加工过程等不确定因素影响。从遗传算法和田口稳健设计的理论方法出发,结合悬臂梁式压电振动换能器的理论模型,设计了以输出平均功率为优化目标的稳健优化设计方案。利用目标函数的梯度信息加速遗传算法的收敛,以遗传算法产生的每一代种群的最优个体作为内表中的一组数据,利用外正交表计算优化目标的信噪比,对优化目标信噪比进行对比分析,得到一组最优设计参数。将该组数据对应的换能器与其它优化方法得到参数对应的换能器性能进行对比,表明该组参数下,换能器有较好的工作特性。     

一种综合的概念语义相似度计算方法

针对概念语义相似度计算的片面性和不完善性问题,该文提出了一种综合的概念语义相似度计算方法。首先,基于传统的语义距离模型,综合考虑了有向边的类型、深度、密度、方向等因素,计算概念相似度;然后,根据本体的五元组组成元素,分别计算概念相似度;最后,使用sigmoid阈值函数求得各个相似度对应的调节因子值,并进行综合,实现了调节因子与各个相似度的自适应。实验证明,该方法有效且精确。     

MATLAB仿真在基带传输教学中的应用

根据基带传输相关课程内容的特点,提出采用MATLAB/Simulink仿真来进行演示和实验以解决相关的问题。同时阐述了使用MATLAB/Simulink对基带传输系统仿真的过程。     

一种提取目标图像最小外接矩形的快速算法

在图像分类与识别算法研究中,目标几何特征的提取通常需要计算目标图像的最小外接矩形以获取长、宽等属性。针对该特点,提出一种利用顶点链码与离散格林理论相结合的方式提取目标图像的最小外接矩形的算法。该算法只需根据顶点链中垂直或水平方向上的点坐标即可求出目标的面积、形心和主轴。基于顶点链码和离散格林的主轴法和旋转法可快速求出目标的最小外接矩形。实验结果表明,旋转法的运算速度是现有算法的2倍左右,主轴法的速度又比旋转法快速2倍左右。     

改进的基于空间特征的划痕检测

改进基于空间特征的划痕检测算法。在延续原有算法的有效性和稳定性的基础上,重新建立划痕空间特征量的阈值选取和宽度自适应参数之间的联系,并添加划痕长度的检测。经实验验证,和原算法相比,改进的基于空间特征的划痕检测算法能更多地检测到视频序列中的划痕,提高了9.2%的准确率,并且功能更加完善,能检测划痕长度。     

基于P-ISOMAP特征脸谱的人脸识别

针对等距映射(Isometric Projection)算法计算量大且易受噪声影响的缺点,根据主成分分析(PCA)兼具降维和除噪的特性,提出一种基于ISOMAP算法的P-ISOMAP特征脸谱(PCA-ISOMAP)人脸识别技术。该算法首先通过主成分分析把人脸图映射到一个人脸图像子空间,然后在人脸图像子空间中进行等距映射降维。P-ISOMAP特征脸谱算法在保持人脸图像的全局几何特性和低维人脸流形的同时,提高了算法效率,能够有效抑制人脸图像中噪声。通过在人脸库中的实验表明,改进后的算法在识别率以及运行效率方面都优于等距映射算法。     

农机政务智能客户端系统研究与实现

为适应行业改革过程中对信息系统需求变化,针对目前政务信息系统基本结构,结合B/S和C/S的优劣,提出了以智能客户端技术为主体,采用工作流、基于角色访问控制,系统安全等技术相结合的敏捷系统框架.以四川农机政务系统的建设过程为实践背景,分析了政务系统的基本组成、系统模型,数据服务框架及业务流程定义、解释和执行机制,用角色控制理论构建直线职能型组织结构中各级角色的访问权和操作权分配控制模型及工作流中各业务活动的权限管理,探讨渗透系统全程的安全机制与实现.     

基于多传感器数据融合的智能小车避障的研究

针对智能小车避障问题,提出了一种将模糊逻辑和神经网络相结合的融合方法—Takagi-Sugeno(T-S)模糊神经网络方法。基于此方法的数据融合算法应用在智能小车避障运动中,采用多只超声波传感器和红外线传感器探测障碍物的距离和方向,采集的各种数据利用T-S模糊神经网络进行融合。通过实验仿真表明:此方法能够使智能小车对障碍物的灵活避障和导航行进。